Skip to main content

Full text of "Psychological anthropology; approaches to culture and personality"

See other formats


Digitized by the Internet Archive 

in 2010 with funding from 

Lyrasis IVIembers and Sloan Foundation 


Approaches to Ctdhi-re and Personality 




U/ih'ersify of Chicago Cornell University 

Argyris Understanding Organizational Behavior 

Adams & Preiss (eds.) Human Organization Research (Pub- 
lished for the Society for Apphed Anthropology) 

Hsu (ed.) Psychological Anthropology: Approaches to Culture 
and Personality 

Psychological Anthropology 

Approaches to Culture and Personality 

Edited by FRANCIS L. K. HSU 

Chairman, Department of Anthropology 
Northwestern University 


Homewood, Illinois • 1961 



First Printing, July, 196 1 

Library of Congress Catalogue Card No. 61-15062 



The purpose of this volume is twofold. On the one hand, it is an 
assessment of the up-to-date gains in the field of culture-and- 
personality. Each of the contributors tries to achieve comprehen- 
siveness within the scope of his particular assignment. Insofar as 
possible each brings together materials from diverse sources, from 
obscure journals to their own yet unpublished field notes. On the 
other hand, each of them also attempts to indicate some of the most 
important problems yet to be tackled. All the contributors outline 
some of these problems, the hypotheses and methods most relevant 
to their investigation, and possible solution. 

The American tradition in textbooks is that they contain mate- 
rials from the beaten paths and are exercises in facts and principles 
generally endorsed by most or all scholars. Such a tradition fails to 
introduce the student to the vitality of an expanding and exciting 
discipline. This book is a textbook, but tliere will be many contro- 
versial spots in it. The reader will find no complete agreement 
among the contributors, nor between the contributors and the 
editor. This is a text in which differences in facts, theories, and 
points of view are not only pointed out, but also explored at some 
length, leading, in some instances, even to almost diametrically 
contrasting conclusions between the authors. 

Another reason for our approach to this text is that, since the 
subdiscipline of culture-and-personality is only about twenty-five 
years old, we are severely limited by the availability of well- 
established facts and principles. If we only aim at the beaten paths, 
then we would have either to confine ourselves to the obvious or 
have little to say. Culture-and-personality has simply not had the 
accumulation of scholarly heritage enjoyed by older subdisciplines 
of the science of man such as archaeology or linguistics. 

These two reasons are interrelated. The paucity in culture-and- 
personality of beaten paths points to the need for growth. And 
growth is impossible without strong efforts to explore new and un- 
sure grounds. In an interdisciplinary subject such as ours, explora- 
tion of new and unsure grounds will by definition be a major part 
of its endeavour for years to come. 

Throughout this enterprise I am fortunate in having a group of 


colleagues as collaborators who have spared no pains in giving me 
their generous support and gracious cooperation. In particular I 
would like to acknowledge my indebtedness to Thomas Gladwin, 
Anthony F. C. Wallace, and Donald T. Campbell, who are also 
contributors to this book, and to Paul J. Bohannan. But my grati- 
tude to all contributors to this volume is very considerable. Their 
intellectual distinction as scholars is a matter of public knowledge 
in the professions and needs no advertising from me. But I, for my 
part, am compelled to express my gratitude to all thirteen con- 
tributors to this volume for their forbearance in the face of my 
many requests, demands, and even, at times, impudence. As a per- 
son born and brought to adolescence in traditional China, I know 
I am open to the suspicion (on the part of those who have read 
Chapter 14) that, still prompted by my early culture pattern of 
mutual dependence, I protest gratitude merely as a matter of good 
form. In the present instance, however, my contributors completely 
and truly deserve my gratitude. As a result of laboring as Editor of 
this volume I have the great satisfaction not only of seeing our joint 
efforts come to fruition, but also of receiving invaluable intellectual 

I wish also to express my indebtedness to Mrs. Elizabeth E. Reed, 
Mrs. Sharon Horine, and Mr. Robert Hunt, who have most ably 
assisted me in the final preparation of the manuscript for publica- 

F. L. K. H. 



I. Psychological Anthropology in the Behavioral Sciences, 

Francis L. K. Hsu i 


Editor's Introduction 17 

2. Japan, Edward Norbeck and George De Vos 19 

3. Africa, Robert A. LeVine 48 

4. North America, John J. Hou'igmann 93 

5. Oceania, Thomas Gladwin 135 

6. National Character and Modern Political Systems, Alex 
Inkeles 172 

7. Am.erican Core Value and National Character, Francis 

L. K. Hsu 209 


Editor's Introduction 231 

8. Cross-Cultural Use of Projective Techniques, Berf Kap- 
lan 235 

9. Mental Illness, Biology, and Culture, Anthony F. C. Wal- 
lace 255 

I o. Anthropological Studies of Dreams, Roy G. D'Andrade . 296 

1 1 . The Mutual Methodological Relevance of Anthropology 

and Psychology, Donald T. Campbell 333 


Editor's Introduction 353 

12. Socialization Process and Personality, John W. M. Whit- 
ing 355 

13. Culture and Socialization, David F. Aberle 381 




14. Kinship and Ways of Life: An Exploration, Francis L. K. 

Hsu 400 


Editor's Introduction 457 

15. An Overview and a Suggested Reorientation, Mel ford E. 
Spiro 459 


A Selected Bibliography Bearing on the Mutual Relationship 

between Anthropology, Psychiatry, and Psychoanalysis 493 

Photographs following 498 


Author Index . 501 

Subject Index 509 


David Aberle, Professor of Anthropology and Chairman of the 
Department, Brandeis University 

Roy D'Andrade, Instructor of Anthropology, School of Educa- 
tion, Harvard University 

Donald T. Campbell, Professor of Psychology, Northwestern 

George DeVos, Associate Professor, School of Social Welfare, 
Associate Research Psychologist, Institute for Human Develop- 
ment, and Research Associate, Center for Japanese Studies, Uni- 
versity of California at Berkeley. 

Thomas Gladwin, Social Science Consultant, National Institute 
of Mental Health 

John J. HonigmAnn, Professor of Anthropology, University of 
North Carolina 

Francis L. K. Hsu, Professor of Anthropology and Chairman of 
the Department, Northwestern University 

Alex Inkeles, Professor of Sociology, Harvard University 

Bert Kaplan, Professor of Psychology, University of Kansas 

Robert A. LeVine, Assistant Professor of Anthropology, Com- 
mittee on Human Development, University of Chicago 

Edward Norbeck, Professor of Anthropology and Chairman of 
the Department, WiUiam Marsh Rice University 

Melford Spiro, Professor of Anthropology, University of Wash- 

Anthony Wallace, Professor of Anthropology and Chairman of 
the Department, University of Pennsylvania 

John W. M. Whiting, Professor of Anthropology, School of 
Education, Harvard University 

chapter i 



Nortknvestern University 

At the beginning of our joint efforts the contributors to this vol- 
ume were of the opinion that attempts at dehneating boundaries for 
culture-and-personahty would do more harm than good. Too often 
precise boundaries have been used as excuse for lack of data, 
methods, and results. What we need in culture-and-personality is 
not orthodoxy but more specific research and discussion. Some opin- 
ions were milder than others but the direction of our comments was 
similar. One commented in the following vein: 

I feel that any area of study which is still as formative as ours can readily deal 
itself out of important areas of inquiry by a premature setting of limits. Anthro- 
pology itself supplies a classic example. The respective areas of study of archaeol- 
ogy, physical anthropology, and ethnology were so neatly defined and separated 
that it took years of effort and the pressure of great intellectual need to recon- 
stitute the connective tissue which had been unthinkingly destroyed by the classi- 
ficatory surgery once fashionable. We are not an exclusive society which needs 
entrance requirements for members. 

Others expressed themselves as follows: 

My advice is not to worry about these demarcation problems but to go where 
our interests and talents lead us. Anthropology has always been distinguished 
by amoeba-like extensions into any discipline where its problems or interests have 
pushed it. 

The virtues of anarchy and chaos are many — the pains of efforts to achieve unity 
are usually without compensatory gain. 

Finally, I do not think we should repeat the errors of many in the standard aca- 
demic disciplines by taking the boundaries of our field too seriously. A recent 
article begins with a comment that is relevant here: "It is perhaps a reflection of 
the intellectual insecurity of social scientists that they spend an inordinate amount 
of time and energy defining the boundaries of their respective fields as if these 
were holy lands which had to be defended against expansive, barbaric, and heathen 


I received more reactions on this subject than I have reproduced 
here. They do not all state the point as strongly but there is none 
which upholds the opposite view. In the circumstances it is natural 
that our ideas as to what should be the proper concern of culture- 
and-personality vary a good deal. As I scan our correspondence and 
the sometimes copious notes of our various meetings, I find the 
following general trends of thought: 

1. A work of culture-and-personality is one by an anthropologist who has a 
good knowledge of psychological concepts or by the member of another discipline 
who has a good knowledge of anthropological concepts. 

2. Any work that deals with the individual as the locus of culture. 

3. Any work that gives serious recognition to culture as an independent or a 
dependent variable associated with personality. 

4. Any work by an anthropologist which uses psychological concepts or tech- 
niques or by a scholar in a psychological discipline which provides directly perti- 
nent data in forms which are usable by anthropologists. 

5. The field of culture-and-personality is equivalent to the cross-cultural study 
of personality and sociocultural systems and includes such problems as (a) the 
relation of social structure and values to modal patterns of child rearing, (b) the 
relation of modal patterns of child rearing to modal personality structure as ex- 
pressed in behavior, (c) the relation of modal personality structure to the role sys- 
tem and projective aspects of culture, and (d) the relation of all of the foregoing 
variables to deviant behavior patterns which vary from one group to another. The 
theories used and hypotheses tested can come from any of the behavioral sciences, 
but the characteristic mark of culture-and-personality research is the emphasis on 
natural group differences as the subject matter. Studies of individual differences 
are not, therefore, works of culture and personality. Nor are studies of the experi- 
mentally produced group differences of many social psychologists. Studies of many 
role personalities within a particular society are on the borderline, but group dif- 
ferences within a society are, in my opinion, squarely within the culture and 
personality field. Thus, Marvin Opler's studies of types of schizophrenia in two 
American ethnic groups are culture and personality work. 

6. The conception of personality-culture as emergent from intereaction is fruit- 
ful. To this it should be added that students of culture-and-personality are con- 
cerned with behavior always with reference to its antecedents and cannot be 
satisfied simply to describe its characteristics — as social psychologists are wont 
to do. 

The possible differences between culture-and-personality and so- 
cial psychology will be touched upon later. Our own lack of agree- 
ment is probably reflective of the perennial seesaw discussion among 
many anthropologists on the same question. At one end is Kroeber's 
concept of the Superorganic, and the perhaps more extreme posi- 
tion of Leslie A. White in his CuUurology, which comes close to as- 
serting that the march of history is independent of the birth of 


particular personalities and that cultures transcend the minds and 
bodies of the individuals living in them (Kroeber 1948:253-255 
and White 1949) . This view has been criticized on the ground that 
culture cannot exist without the individual, since "to objectify a 
phenomenon that can have no manifestation except in human 
thought and action is to argue a separate existence for something 
that actually exists only in the mind of the student" (Herskovits 
1948 : 25) . At the other end are students who note individual differ- 
ences and cultural variation in each society. Herskovits shows how 
the same song, "John Crow," prevalent in the northwestern part 
of Jamaica, is rendered into many different versions by many dif- 
ferent singers (Herskovits 1948:565-569). John Gillin demon- 
strates the intrasocietal differences among many nonliterate socie- 
ties (Gillin 1939:681-702). Bert Kaplan has revealed similar 
differences in four American Indian cultures (Kaplan 1954) . Hart 
perhaps pushed the importance of individual personality differences 
to a greater extent than most others (Hart 1954) . 

Needless to say, Kroeber was not unaware of the fact that cul- 
tures have to be expressed through individuals, for in his major text- 
book he did devote a whole chapter to "Cultural Psychology" 
(1948:572—621) . On the other hand, it is equally obvious that no 
science of man is possible if we merely concentrate on individual 
differences. Perhaps it is to satisfy both extremes that Kluckhohn 
and Mowrer found it desirable to introduce an extensive analysis of 
all the components of personality, from the biological, the physical- 
environmental, the social, and the cultural, on the one hand, and the 
universal, the communal, the role, and the idiosyncratic on the 
other (Kluckhohn and Mowrer 1944:4) . 

In the 1948 edition of their anthology entitled Personality in Na- 
ture, Society and Culture, Kluckhohn and Murray reformulated 
the four determinants which were then designated as constitutional, 
group membership, role, and situational. The two editors con- 
cluded in the second edition of this book that "the differences 
observed in the personalities of human beings are due to varia- 
tions in their biological equipment and in the total environment 
to which they must adjust, while the similarities are ascribable to 
biological-environmental regularities" (Kluckhohn and Murray 

In a later publication Kluckhohn leaned more strongly toward 
the cultural factor in human behavior, but modestly guarded him- 
self as follows: 


In all these, and other, secondary categories the basic determinants are con- 
founded — as the statisticians would say or, in other words, behavioral scientists 
must deal with a complex field structure. There is, at best, a vague recognition 
that all are involved, but, in practice, scientists from different disciplines and with 
different temperamental biases tend to operate as if motivation were, after all, 
simply biological or situational or cultural. We lack the techniques, quantitative 
or otherwise, for dealing with systems of organized complexity. And so, for the 
time being at least, we must do the best we can with crude, first approximations. 
Each of us must continue to insist that the particular variable he is most interested 
in be taken fully into account. If there is a reasonable of "give" on every side, if 
each specialist fully accepts the fact that his discipline can explain not everything 
but something, the results are not too bad. One may compare a game in which 
the high card or combination is crucial. Other cards in the hand have a value but 
a secondary importance for that deal. Some hands are dealt by science where the 
winning combination is certainly held by biology, others where psychology, so- 
ciology, geography, or anthropology can do the calling. So I shall here unashamedly 
concentrate upon the hands where, it seems to me, anthropology can bet high 
upon the significance of cultural factors for understanding and explanation. 
(Kluckhohn 1954:3—4) 

The Core of Culture-ond-Personality 

Though as a group we eschew boundaries, I think it is quite ap- 
propriate for the editor at least to offer some thoughts on the cen- 
tral concerns of culture-and-personality. In this venture I do not 
expect to settle anything. In such a fuzzy jirea no clarification is 
likely to meet with universal acceptance. What I shall try to do is 
not more than to offer some material to feed further discussion. 

It is probably as trite to observe that all human behavior is medi- 
ated through the minds of individual human beings as it is to observe 
that all human individuals live in social groups each governed by a 
specific pattern of culture. All human behavior, except random 
movements and reflexes, is, therefore, at once psychological and so- 
cial in nature. However, the same psychosocial data may be ap- 
proached from different angles. The angle of approach would seem 
to be the primary difference between cultural anthropology and 
social anthropology, and between them and culture and personality. 

Social anthropology began in Britain. The British view is that 
social anthropology is synonymous with anthropology and sociology 
combined. That is to say it deals with all aspects of human behavior 
from kinship and political organizations to economies and religions. 
The American definition is that social anthropology is confined to 
the study of social or political organizations. Therefore, some 
American students are surprised that E. E. Evans-Pritchard, a well- 


known social anthropologist, should have "left" his field to write a 
book on Naer B^eligion (1954). There are perhaps two ways of 
seeing the real difference between them. First, the British social 
anthropologists really have distinguished themselves in the inten- 
sity of their field work and of analysis of the data, a trend first begun 
with Rivers and Williams in the Torres Straits and later Malinowski 
in the Trobriands. Thus, Evans-Pritchard carried out his field work 
in the late twenties and spent the next thirty years publishing pri- 
marily on the two societies he studied. This is more or less true of 
other well-known British social anthropologists such as M. Fortes, 
R. Firth, and Max Gluckman, though most of them have studied 
more than one people. This contrasts with the less intensive pattern 
of field work among American anthropologists, characterized by 
shorter periods of sojourn, lack of emphasis on thorough familiarity 
with the native language, and even among many students of the 
Historical School, a relatively greater emphasis on problem orienta- 

The other way is to see cultural anthropology as largely dealing 
with human behavior in terms of products (culture traits, rituals, 
dances, techniques, and so forth) while social anthropology, in 
terms of relationships (such as kinship, inheritance, law, and gov- 
ernment). According to this view cultural anthropology studies 
the end results — cultures, including their diffusion from area to 
area and their development from epoch to epoch; and social anthro- 
pology studies the interpersonal mechanisms through which human 
beings learn, manipulate, and produce cultures. 

Neither of these distinctions is complete in itself. The British and 
American patterns of field work reflect no real difference in scope, 
only, with many obvious exceptions and to a certain extent, in 
thoroughness and depth. They are, in fact, complementary ap- 
proaches to the same objective. The British way often leads the 
field worker into a displaced ethnocentrism in which Bongo ethno- 
centrism takes the place of English ethnocentrism. The American 
way sometimes leaves the field worker with many factual details 
but with possibly less sensitivity to the feelings and the views of the 
peoples he has studied. The product versus relationship distinction 
is equally without finality. For as the student intensifies his re- 
searches, his concern for the mechanisms will inevitably lead him 
to the end results and vice versa. Many anthropologists will prob- 
ably deny the existence of either of these distinctions. 

It is in this context that we must view culture-and-personality. 


Culture-and-personality deals with human behavior primarily in 
terms of the ideas which form the basis of the interrelationship be- 
tween the individual and his society. On the one hand, it deals with 
ideas shared by a considerable portion of any society: the "shame" 
or "guilt" feelings among the Japanese, the belief in immanent jus- 
tice among some children of Ghana, the anxiety about food in ex- 
cess of the actual danger of going hungry among some Oceanic 
peoples, and even the world view of the Chinese; how these and 
other ideas held by the individuals are rooted in the diverse patterns 
of culture in which they grow up. On the other hand, culture-and- 
personality deals with characteristics of societies: reactions to con- 
quest and disaster, internal or external impetuses to change, mili- 
tarism and pacificism, democratic or authoritarian character; it 
deals with how these and other characteristics consistently associ- 
ated with some societies may be related to such things as the aspira- 
tions, fears, and values held by a majority of the individuals in these 

With these thoughts on the central concerns of culture-and-per- 
sonality, I would like to propose a new title for our subdiscipline: 
psychological anthropology. 

For over twenty years culture-and-personality has retained its 
cumbersome title. I think the time has probably come for us to give 
it a less cumbersome and more logical title. The concept of person- 
ality, which anthropologists have borrowed from psychologists, 
leads to some difficulties. For example, some anthropologists, 
though resorting to psychological explanations at many crucial 
points of their arguments, tend to regard the personality concept 
either as indistinguishable from culture or as much deeper than 
what the anthropologist can usually deal with. In his book The 
Foundations of Social Anthropology Nadel expresses the following 

We may take it for granted that there is some connection between the make-up 
of a culture and the particular personality (or personalities) of its human carriers. 
Yet in taking this connection to be a simple and obvious one, so simple and obvious 
that one can be inferred from the other, v/e run the risk of arguing in a circle and 
of using the word "personahty" in an ambiguous sense. For by "personality" we 
can mean two things. We can mean, first, the sum-total of the overt modes of 
behaviour of an individual, in which we discern some integration and consistence, 
and which we thus understand to be facets or "traits" of that total, patterned 
entity. Or secondly, we can mean some basic mental make-up underlying the pat- 
tern of overt behaviour and accounting for it in the sense of a "hidden machine" 
or a causally effective set of factors. (Nadel 1951:405) 


Nadel then refers to a distinction made by the psychologist, R. B. 
Cattell, between "surface traits" "which give us 'clusters' of in- 
trinsically related characteristics of behavior observable in every- 
day Hfe," and "source traits" or "factors," "which are extricated 
by analysis and have a causal significance, being possible explana- 
tions of how the actually existing cluster forms may have origi- 
nated" (Cattell 1946:4). Nadel's reasoning goes as follows: If the 
anthropologist operates with the personality concept and wishes to 
ascertain the mental make-up of a group possessing a certain cul- 
ture, he should resort only to tests and other techniques developed 
by psychology. If he wishes to "define the cultural patterns in terms 
of 'basic' psychological agencies," he "must examine them where 
they are ultimately rooted — in the individual" (Nadel 195 1 1407) . 
But if the anthropologist approaches the personality merely from 
cultural observation, by means of direct inference, he can only 
reach the "surface traits." Even though he may infer the desires, 
motivations, and so forth, prompting the overt behavior he "pene- 
trates, as it were, only a short distance beneath the surface;" for 
the desires and so forth are "simply implicit in the cultural mode of 
behavior," or "are merely its sustaining energies, and have no causal 
and explanatory significance" (Nadel 1951:405). Nadel con- 

As long as we are inferring personality types from cultural observation we can- 
not legitimately claim any explanatory value for the personality concept; if we 
did, we should be committing the cardinal sin in science, namely, of pronouncing 
upon invariant relations between facts which are not "demonstrably separate." 
(Nadel 1951:407) 

I think Nadel is wrong here. Psychological constructs, by virtue 
of the fact that they have to be inferred from linguistic data or 
other indirect evidences supplied by the actors, are certainly 
"demonstrably separate" from behavior which can be directly ob- 
served. Furthermore, gravitation is inferred from falling apples, 
rises and falls of tides, and movements of the moon, earth, and other 
heavenly bodies. Gravitation can never be seen anywhere except in 
terms of what it does, through the behavior of the objects which it 
controls or influences. Similarly physical hunger can only be in- 
ferred by stomach contractions, nausea (if the hunger is severe 
enough) , or malnutrition of the body (as a result of prolonged 
hunger) . No one can see hunger except through these and other 
concrete expressions of it. I have yet to hear from a scientist who 
denies the usefulness of the concept of gravitation or hunger, and 


who insists that correlating certain movements of the heavenly 
bodies with gravitation, or correlating certain physiological phe- 
nomena with hunger, is equivalent to commiting "the cardinal sin 
in science, namely, of pronouncing upon invariant relations be- 
tween facts which are not demonstrably separate." As our knowl- 
edge progresses, we may conclude that the concept of gravitation 
or hunger is no longer adequate to account for certain phenomena, 
but we cannot deny that during a certain period of our scientific 
development these concepts have played crucial and organizing 

However, Nadel's arguments do point up one important matter, 
namely the personality which psychological anthropologists deal 
with is not the same as that of the individual psychologists. At 
least conceptually, the latter deal with the unique personality of 
the individual, but the former deal only with those character- 
istics of the individual's mind which are shared as part of a wider 
fabric of human minds. In Chapter 8 Kaplan discusses various at- 
tempts to conceptualize and understand the differences between 
the two kinds of reality, for example, by introducing the term "so- 
cially required" personality patterns as distinguished from the "ac- 
tual modal" personality patterns. Yet the term "personality" 
possesses connotations that often lead the student to regard it as 
a complete entity in itself. Instead of seeing personality as a life- 
long process of interaction between the individual and his society 
and culture, he thinks of it as being some sort of reified end- 
product (of very early experiences according to orthodox Freu- 
dians, of somewhat later sociocultural forces according to many 
Neo-Freudians and social scientists), which is ready to act in this 
or that direction regardless of the sociocultural fields in which it 
has to operate continuously. It is true that the scholars have never 
quite said so in exact words. It is also true that "field theory" of 
Kurt Lewin or others has many advocates. But given the social 
scientist's individualist culture heritage of hero and martyr wor- 
ship, and a Judaeo-Christian theological background of absolute 
conversion and final salvation, the one-sided finished-product view 
of personality would seem too "natural." Such a view must be re- 
sisted and the beginning of such a step is to eliminate the word per- 
sonality from the title of our subdiscipline. 

Some anthropologists may object to the new title of psychologi- 
cal anthropology on several grounds, though I see no insurmount- 
able obstacles against it. One argument is that it may lead to pro- 


liferation of subdisciplines. But giving the subdiscipline a more 
logical name should not cause any more proliferation than culture- 
and-personahty has done. In the second place, division of any large 
single discipline into subdisciplines is inevitable as our knowledge 
in that area grows. Seventy-five years or so ago it was as sufficient 
simply to be an anthropologist as it was to be a sinologist. But soon 
anthropology was divided into cultural anthropology, physical 
anthropology, and so forth, and we no longer find the term sinolo- 
gist except in some ultraconservative academic pockets. The same 
phenomenon has occurred in biology, physics, chemistry, and even 
such subdisciplines as linguistics and geometry. The only caution 
that we must exercise in branching out is that we must make sure 
that the advances of knowledge are ahead of the subdivision and 
not vice versa. 

Another argument against the new title is that it will turn out 
to be neither psychology nor anthropology, that it is a no man's 
land. This is not a fruitful argument. We have textbooks on physio- 
logical psychology, biochemistry, astrophysics, and psychosomatic 
medicine. There is not the slightest indication that the separate dis- 
ciplines which have been so allied with each other have suffered in- 
tellectually. On the contrary psychosomatic medicine has enriched 
both psychology and medicine; and, without biochemistry, biology 
and chemistry would both have been poorer. The psychological 
anthropologist should certainly make use of the results not only in 
psychology but also in psychoanalysis, sociology, and even experi- 
mental psychology and philosophy wherever these are relevant and 
applicable. This is the way all sciences grow, like so many amoebae 
which extend a pseudopodium here and another there, retracting 
them here and there while their nuclei remain more or less constant. 

Psychological Anthropology and Related Disciplines 

To clarify our thoughts further, it may be advantageous to ex- 
amine the relationship between psychological anthropology and a 
few other disciplines. In the short history of psychological anthro- 
pology as a subdiscipline the clinical sciences have figured largely. 
In fact the indebtedness of psychological anthropology to psychia- 
try and psychoanalysis is immeasurable. Anyone who knows any- 
thing about psychological anthropology can easily call to mind the 
significant roles of such clinicians as Abram Kardiner, Erik Erick- 
son, Alexander Leighton, Karen Horney, Erich Fromm, Geza 
Roheim and associates, and of course the master himself, Sigmund 


Freud. These students have, either singly or in collaboration with 
anthropologists, helped to make psychological anthropology grow 
immensely in stature, concepts, and volume of research. 

However, psychological anthropology is not a clinical science 
and, while it has benefited from the clinical sciences, it has its own 
ways. In objectives, psychological anthropology is concerned with 
large numbers of individuals who are normal and functioning 
members of their societies. In methods of approach psychological 
anthropology follows the usually accepted scientific procedure of 
hypothesis formation, testing of hypothesis, cross-cultural valida- 
tion of the result, and further refinement of the hypothesis. In this 
it must first be emphasized that psychological anthropology is not 
simply the psychology of the individual, and it must shun psycho- 
analysis of whole cultures in the manner that Freud arrived at his 
conclusion on the origin of totem and taboo (Freud 1919) . 

What psychological anthropology deals with are (a) the con- 
scious or unconscious ideas shared by a majority of individuals in 
a given society as individuals (which can be subsumed under such 
terms as basic personality or modal personality (Linton 1945 : 1 30) , 
both statistical or nearly statistical concepts) and (b) the con- 
scious or unconscious ideas governing the action of many indi- 
viduals in a given society as a group (sometimes described as group 
psychology, mob psychology, or collective conscience) . Both of 
these are different from the unique psychology of the individual. 
It is not maintained that the ideas underlying the life pattern of a 
group and those of the actions of an individual are two distinct 
entities. In fact they form a continuum. There is much evidence 
to indicate that many individuals evaluate national or international 
affairs in terms of their own personal likes and dislikes, anxieties, or 
aspirations. But before the psychological anthropologist can con- 
clude that one is rooted in the other, he must make sure that he 
is not arguing merely from analogy, that he has made sure he is not 
confusing broad trends of cultural development, which may be 
psychologically propelled, with specific institutional details, which 
are usually historically determined. 

The damage to psychological anthropology by the failure to dif- 
ferentiate the normal from the abnormal is great. Admittedly the 
demarcation line between the normal and the abnormal is not clear. 
Nevertheless, even after allowing cultural differences, there is still 
undeniable evidence for certain core differences between them 
(Hsu 1952:238-248). The extension of the abnormal psychology 


of the individual into the normal pattern of the group was, I think, 
responsible for Freud's lopsided emphasis on the death instinct, and 
its continued lopsided emphasis by modern Freudians of many hues. 
That all human beings die is as indisputable as the fact that all 
human beings live. The extent to which some human beings ap- 
parently seek self-destruction in wars, suicides, alcoholism, and 
psychotic behavior, and the possible psychological mechanisms un- 
derlying such patterns of action have been brilliantly outlined by 
Jules Masserman (1955: 647-649 ) . 

However, while the evidence in support of the universality of the 
life instinct among humans and animals is overwhelming, the evi- 
dence in support of the death "instinct" comes chiefly from the 
relatively abnormal. This is why the self-destructive tendencies 
are not common among the majority of any society. Furthermore, 
the incidence of suicide and homicide no less than delinquency and 
adventure vary from culture to culture. Clearly another type of 
explanation than a universal postulate of death instinct is indicated. 
It will probably be well for psychological anthropologists to be on 
guard against generalizing from the psychology of a minority of 
the relatively abnormal to that of a majority of the relatively nor- 
mal when they make use of the psychiatrically derived resources, 
insights, and data. 

Among all the behavioral sciences, psychological anthropology 
and social psychology have the future potentiality of developing 
the closest and most mutually enriching relationship with each 
other. Both disciplines deal with society and both deal with psychol- 
ogy, but they have been separated from each other so far in sig- 
nificant ways. We have seen that two of the points made in the 
preliminary discussions among the contributors to this volume 
were: (a) that the characteristic mark of culture-and-personality 
research is the emphasis on natural group differences along ethnic or 
societal lines, and so forth, as the subject matter, whereas social 
psychology often deals with experimentally produced group dif- 
ferences; and (b) that culture-and-personality scholars are con- 
cerned with behavior always with reference to its antecedents, while 
social psychologists are satisfied simply to describe its characteris- 
tics. I do not think the second distinction to be valid, for many 
studies in social psychology are attempts to discover the antecedents 
of behavior; and I think the first distinction is only partially valid, 
since psychological characteristics due to role, sex, and occupational 
affiliations are also problems of psychological anthropology. 


What have so far differentiated psychological anthropologists 
from the social psychologists are found in three areas. First, psycho- 
logical anthropology is cross-cultural in approach from its incep- 
tion while social psychology has traditionally drawn its data from 
Western societies. Second, social psychology is quantitative and 
even, to a certain extent, experimental in orientation, while psycho- 
logical anthropology has paid little attention to research designs 
and only lately awakened to the need for rigor in the matter of 
hypothesis formation and of verification. 

In both of these connections the distance between the two dis- 
ciplines is narrowing, and rightly so. Social psychologists have be- 
come increasingly more interested in cross-cultural validity of their 
generalizations. This anthropological contribution to psychology is 
well recognized by Campbell, a social psychologist, in Chapter 1 1 
of this volume. A comparison of the earlier and later editions of 
many texts on social psychology shows far greater use of cross-cul- 
tural data in the later than in the earlier works, though some such 
as Klineberg (1940 and 1954) have always led among the pioneers 
in interdisciplinary thinking and research, while others such as 
Bogardus (1950) are less inclined in that direction. In fact, it is a 
rare textbook on social psychology today which does not contain at 
least references to Margaret Mead, Ruth Benedict, Geoffrey Gorer, 
Clyde Kluckhohn, Ralph Linton, M. J. Herskovits, John Whiting, 
or some other anthropologists. Psychological anthropologists, on 
their part, have become increasingly more sensitive to the impor- 
tance of sophistication in research designs and quantification. The 
chapters by Wallace, Whiting, Aberle, Spiro, and D'Andrade in this 
volume and the works of Hallowell, Kluckhohn, Gillin, and others 
are, in different ways, objective evidence in this new direction. The 
psychological anthropologist may not agree with (or may not be 
able to do much about it at the moment even if he does agree with) 
some of the methodological points raised by Campbell in Chapter 
1 1 , but there is no doubt about the importance of such thinking 
to psychological anthropology. Psychological anthropology has 
already derived no small part of its methodological inspiration from 
social psychology and, as time goes on, its indebtedness to social 
psychology is likely to be even greater than its previous indebted- 
ness to the clinical disciplines. 

The third area in which psychological anthropology differs from 
social psychology thus far is that it deals not only with the effect of 
society and culture on personality (a basic concern of social psy- 


chology) but also with the role of personality characteristics in the 
development, formation, and change of culture and society. Chap- 
ters 7, 8, 12, 13, and 14 of the present volume touch upon this in 
different ways. Finally, in Chapter 15 the reader will find a hy- 
pothesis to investigate the mechanism underlying the mutual in- 
fluences between the individual society and culture. For a sound 
theory which aims at explaining the relationship between man and 
culture must not only account for the origin of psychological char- 
acteristics as they are molded by the patterns of child rearing, social 
institutions, and ideologies but must also account for the origin, de- 
velopment, and change in these child-rearing practices, institutions, 
and ideologies. It is a well-known fact that societies and cultures do 
change, often slowly but sometimes drastically. Since human be- 
ings are not so many helpless creatures simply being pushed by ex- 
ternal forces such as geographical calamities, foreign conquests, 
fate, gods, or the unaccountable vicissitudes of some superorganic, 
we must at least find part of the explanations for cultural and so- 
cial changes in the interaction between the human minds and the 
societies and cultures in which they operate. 

At the beginning of this attempt at clarifying our thoughts on 
psychological anthropology, I noted the difficulties besetting such a 
venture. What I hoped to do was not to close the discussion but to 
keep it going. Furthermore, just as a mere matter of emphasis or 
point of view separates cultural anthropology from social anthro- 
pology, so psychological anthropology is similarly differentiated 
from its related disciplines. For example, a cultural anthropologist 
will ultimately come to analyze the ideas behind the diffusion of cer- 
tain cultural traits and complexes; a social anthropologist will ulti- 
mately look at the material wealth involved in the different forms of 
social organization, exactly as the psychological anthropologist will 
ultimately relate the conscious or unconscious ideas to both par- 
ticular cultural end results and particular human relationships. It 
is probably desirable, however, for the student from one viewpoint 
to hold on to his particular viewpoint as he probes deeper and deeper 
into his data, or else he may be hopelessly enmeshed in them without 
guideposts to go forward or backward. The significance of such a 
viewpoint to the field worker is comparable to that of the "ego" 
to the maker of a kinship chart. As the maker of a kinship chart 
cannot change the "ego" in it without getting lost, the field worker 


who shifts from one viewpoint to another, or has no viewpoint at 
all, is hkely to bring back Httle that is of coherent significance. 



1950 Fundamentals of social psychology. New York, Appleton-Century- 

Cattell, R. B. 

1946 Description and measurement of personality. Yonkers, N.Y., World 
Book Co. 

Evans-Pritchard, E. E. 

1954 Nuer religion. London, Oxford University Press. 

Freud, Sigmund 

19 19 Totem and taboo. London (reprinted in The Basic Writings of Sigmund 
Freud, 1938, The Modern Library, New York, Random House) . 

GiLLiN, John 

1939 Personality in preliterate societies. American Sociological Review 4:681- 

Hart, C. W. M. 

1954 The sons of Turimpi. American Anthropologist 54, 2, Part I, 242—261. 

Herskovits, M. J. 

1948 Man and his works. New York, Alfred Knopf. 

Hsu, F. L. K. 

1952 Anthropology or psychiatry: A definition of objectives and their im- 
plications. Southwestern Journal of Anthropology 8:227-250. 

Kaplan, Bert 

1954 A study of Rorschach responses in four cultures. Papers of Peabody 
Museum of Archaeology and Ethnology, Harvard University, Vol. 42, 
No. 2. 
Klineberg, Otto 

1954 Social psychology. New York, Henry Holt (ist ed., 1940). 

Kluckhohn, Clyde and O. H. Mowrer 

1944 Culture and personality: A conceptual scheme. American Anthropol- 
ogist 46:4. 
Kluckhohn, Clyde, Henry A. Murray, and David Schneider 

1953 Personahty in nature, society, and culture. 2d ed.. New York, Alfred 

Kluckhohn, Clyde 

1954 Culture and behavior. In handbook of social psychology, Gardner 
Lindzey (ed.), Cambridge, Mass., Addison-Wesley Press. 

Kroeber, a. L. 

1948 Anthropology. New York, Harcourt Brace & Co. 


Linton, R. 

1945 Cultural background of personality. New York, D. Appleton Cen- 
tury Co. 

Masserman, Jules 

1955 Dynamic psychiatry. Philadelphia, W. B. Saunders Co, 
Nadel, S. F. 

195 1 Foundations of social anthropology, Glencoe, 111., Free Press. 
White, Leslie A, 

1949 The science of culture. New York, Farrar, Strauss & Co. 



The treatment of the subject of psychological anthropology by 
area presents some difficulties. Culture and personality differences 
between tribal groups and national groups within these large areas 
are sometimes so great that the contributors will either have to gen- 
eralize on a relatively superficial level or else have to confine them- 
selves to a few selected studies which already possess intensity and 

There is no adequate answer to these difficulties. In a work of this 
scope it is simply not possible to gain the intensity and depth attain- 
able in a field report on a single tribe or community. A general 
picture of the psychology of a region like North America or even 
Japan is bound to contain fewer details than a monograph on the 
culture and personality of Polish peasants inhabiting a village in 
Ruthenia. The contributors themselves are keenly aware of the dan- 
ger of overgeneralization, or generalization based on scanty data. 
LeVine has indicated, with reference to Africa, some of the clearest 
instances of fallacies resulting from such procedures; Honigmann 
has assembled a fine array of studies among the North American In- 
dians in which more refined designs and techniques have yielded 
composite psychological characteristics of peoples each scattered 
over a large area that bear out the purely qualitative and inferential 
pictures arrived at years earlier. There are many social and cultural 
mechanisms which, on closer inspection, make for psychological 
standardization of large communities. Communal, tribal, or na- 
tional myths are some of these. Communication and diffusion proc- 
esses are others. 

However, even in this section of the book, our interest is only 
partially areal. The areal arrangement is convenient in providing 
the reader with a panorama of the most significant works of psycho- 
logical anthropology in the area of his curiosity. But a problem 
orientation is present in this section of the book, as in the subsequent 
sections. Norbeck and DeVos discuss personality factors affecting 
differential Japanese acculturation on different continents. LeVine 
summarizes problems of infant experiences and the family environ- 



ment; psychocultural interpretation of ritual, witchcraft, and 
dreams; and the problem of differential incidence and types of men- 
tal illness. Honigmann treats the problems of values and of model 
personality. Gladwin analyses the contributions of Mead and co- 
workers and their use of a broadly framed learning theory, and of 
the Kardiner-Linton group and their use of a revised Freudian 
psychology. In these and other materials the problem-minded reader 
will find much that is informative and stimulating. 

The last two chapters analyze the composite psychological char- 
acteristics of some large, modern, and complex societies, relating the 
individual aspirations to the over-all thought and action patterns 
of each group. On the one hand, they indicate fresh approaches 
to the question of generalization on huge societies. Too often na- 
tional character studies have been attacked on a priori grounds, that 
it is "impossible" to gauge the psychological characteristics under- 
lying complex civilizations, with hundreds of millions of individuals 
living in them. But the basic problem is surely one of level of gen- 
eralization. If we look for individual differences, there is no shortage 
of data which compel us to observe that no two individuals are 
identical. But if we raise our sights to a different level, we shall at 
once see that millions of human beings interact with each other, 
voluntarily or involuntarily, in any large society on any one day, 
often sight unseen, apparently without any. significant difficulties. 
This relatively smooth interaction among strangers in any large, 
modern society is a remarkable reality, which will be impossible 
without some high degree of uniformity, not merely in externally 
visible laws, customs, procedures, and usages but also in externally 
invisible ideas, emotions, expectations, and faiths. Yet these two 
analyses arrive at very different conclusions. Is this due to the differ- 
ences in point of view of the two authors? Is this due to differences 
in the kinds of fact upon which the two authors based their gen- 
eralizations? Are the two papers products of different levels of 
abstraction? Or are they evidence that we must employ more pre- 
cise methods? 

Chapter 2 



William Marsh Rice University, and 


University of California 

The objectives of this chapter are to review research in the field of 
Japanese culture and personality, and to appraise it from the stand- 
point of the contribution it has made to theory and the promise of 
future contribution that it holds. 

Anthropological interest in Japan and the Japanese is old, but 
until World War II it was left principally to native Japanese 
scholars, whose publications rarely reached the Western world. For 
decades before the war, Western writers had made many impres- 
sionistic observations on the character of the Japanese, but writings 
on this subject by scholars trained in psychology, sociology, and 
anthropology are principally postwar. Entry of the United States 
into the war served in several ways to direct the attention of Amer- 
ican social scientists to Japanese culture, and it is during the w;ar 
years that research on Japan using modern techniques of person- 
ality-and culture-began, principally under the sponsorship of the 
United States government. The first published studies are papers 
byLaBarre (i945),Gorer (1942, 1943), and others which attempt 
to describe the Japanese personality and relate it to cultural insti- 
tutions of child training. As is well known, Ruth Benedict's The 
ChrysanthemUfn and the Sword also sprang from research con- 
ducted during the war under governmental subsidy. 

Since the end of the war, the number of American social scien- 
tists engaged in research on Japan has grown, and scholarly writings 
on Japanese culture have increased greatly. We no longer regard 
Embree's Suye Miira as modal for Japanese communities. As a re- 
sult of sociological and ethnological research, we have become aware 
of many regional distinctions in Japanese culture and differences 



along lines of occupation and social class. We have also become in- 
creasingly aware that Japanese culture is in a state of rapid transi- 
tion so that observations made at one point in time are often quickly 

Research on Japan concerned with the relationship between per- 
sonality and culture began with American scholars. Since the end 
of World War II, Japanese scholars have also engaged in research in 
this field on their own culture, and interest in the subject among 
native scholars is growing. Publication of a Japanese journal con- 
cerned with studies using projective techniques (Japanese Journal 
of Projective Techniques) began in 1954. A society composed of 
approximately 30 psychologists and anthropologists called Nihon 
Bunka to Nihonjin no Shinrigakuteki Kenkyii no Kai (Society for 
the Psychological Study of Japanese Culture and the Japanese Peo- 
ple) was formed in 1958. During the past decade many relevant 
publications in the Japanese language have appeared, and several 
research projects concerned with Japanese personality and culture 
are now being conducted by Japanese social scientists. The com- 
bined research of native and foreign scholars makes a surprisingly 
large total, and Japan is probably unique in the field of culture- 
and-personality in being the focus of fairly extensive study by both 
natives and foreigners. 

We shall review both Western and Japanese research that has 
been completed and discuss projects now under way. As a mat- 
ter of convenience we shall classify these studies under five major 
headings that are not mutually exclusive: 

1. Broad Approaches to Understanding National Character 

2. Content Analyses of Forms of Expressive Behavior 

3. Studies Using Projective Techniques 

4. Studies of Early Socialization Patterns 

5. Studies of the Japanese Overseas 

Judgment as to the kind of research and the specific studies to in- 
clude has, of course, been in part arbitrary. We have not limited 
ourselves to research conducted by anthropologists, but have in- 
cluded publications in social, clinical, and child psychology, in 
psychiatry, and in other fields when these studies have dealt with 
questions of the relationship between culture and personality. No 
attempt will be made to review all publications relevant to Japanese 
culture and personality. Many publications, especially in the fields 
of psychology and psychiatry, have been omitted or mentioned 
only in passing because they make no attempt to relate traits of 


personality to cultural determinants. For lack of space, a very large 
group of studies of Japanese culture prepared by ethnologists, so- 
ciologists, historians, economists, and political scientists, both Japa- 
nese and Western, are not discussed. Omission is made with full 
awareness that these publications are relevant to an understanding 
of Japanese culture and personality, as they provide vital informa- 
tion on such matters as differences in culture by region and class 
and trends of cultural change. 

Studies of National Character 

Ruth Benedict's The Chrysanthemum and the Sword is probably 
the only major Western publication on Japan that consensus would 
place under the heading of studies of national character. The impact 
of this work upon both the scholarly world and the general public 
of Japan was surprisingly great. Translated into Japanese, it was 
widely read and served as a strong stimulus to Japanese scholarly in- 
terest in personahty and culture. An indication of the importance 
of Benedict's work is afforded by the fact that it served as the topic 
for a series of seminars, well publicized in scholarly circles, in which 
prominent Japanese scholars participated. A summary of the Japa- 
nese critique of Benedict's methodology and conclusions has been 
given us by John Bennett and Michio Nagai (1953) , and we shall 
mention here only the chief criticism that her study presents a static 
picture of ideal upper class patterns of a time gone by, and ignores 
distinctions by social class and changes through time. Jean Stoetzel's 
postwar study. Without the Chrysanthem^um and the Stvord, points 
up change and indicates that much of what Benedict describes does 
not apply to the modern youth of Japan. Benedict's work not only 
stimulated interest in Japanese character but also led to field re- 
search by Japanese scholars, notably T. Kawashima (1951), on 
modes and differences in conceptions of the values in interpersonal 
relations {chit, on, girt) with which her study had dealt. This re- 
search summarizes interviews with country people showing that 
these cultural ideals are less strongly held by them, especially by 
young people, than Benedict reports. 

Benedict's attempt to delineate Japanese character stands out also 
from the standpoint of methodology. As one of the pioneer studies 
of "culture at a distance," it points out the potentialities of this ap- 
proach. As a result of subsequent field research in Japan and the 
critique of the study by Japanese scholars, we are given a better idea 
of its limitations. 


Since the publication of The Chrysanthemtim and the Sword, the 
most extensive research aimed at understanding Japanese national 
character has been the interdisciplinary studies conducted from 
1953 to 1955 by the Human Relations Research Group, headed by 
Tsuneo Muramatsu, Professor of Psychiatry at Nagoya National 
University. Included in the group were Japanese scholars from vari- 
ous disciplines and one American, George De Vos. The principal 
objective of this project was to determine both modes and regional 
differences in cultural values as these are related to types of per- 
sonality. With the efforts of as many as 30 researchers, samples were 
taken of attitudes and customs of urban and rural populations of 
central and southwestern Japan. Data were also gathered on social 
and economic backgrounds, and a number of families selected as 
modal were subjected to intensive interviews. Principal test instru- 
ments used were the F Scale, derived from American studies of the 
authoritarian personality; two opinion scales devised to test atti- 
tudes toward familial relations and "liberal-traditional" attitudes 
toward Japanese values; the Rorschach test; the Thematic Apper- 
ception Test modified for Japanese culture; a problem situation 
test; figure drawings; a "child-parent problem" test, and question- 
naires on customs of child training. Photographs were also taken 
to illustrate mother-child relations during the first few years of life. 

Samples totaling 250 individuals were obtained from three rural 
settlements, a mountain community depending for subsistence on 
farming and forestry, a fishing community, and a lowland rice- 
farming community, which represent the spectrum of the con- 
ventional scholarly Japanese classification of types of rural 

A sample of over 2,000 individuals was obtained in the cities of 
Nagoya and Okayama, although not all persons of this group were 
subjected to the entire battery of tests. Data gathered under this 
research project are still in the process of analysis and interpretation. 
Although no generalizations concerning modal traits of the Japa- 
nese personality have as yet emerged, a series of publications pre- 
senting interpretations of smaller scope have appeared or are now 
in press. ^ Results of these studies will be discussed in this paper. 

A second major research project aimed at determining regional 
variations in traits of personality and the cultural factors which 
have brought them into existence is now in progress under the di- 
rection of Seiichi Izumi of Tokyo University. This project is also 

^ These include articles by Marui, Murakami, De Vos, and Wagatsuma, cited in bibliography. 


interdisciplinary. It makes extensive use of projective tests and in- 
cludes among its objectives an assessment of national character. Re- 
search is centered upon northeastern Japan and other areas which 
have not previously been subjected to intensive investigation using 
the techniques of culture-and-personality. 

A number of studies by Japanese scholars, some of which are dis- 
cussed below, touch in varying degree upon Japanese national char- 
acter. A recent book (Sofue and Wagatsuma 1959), based upon 
Benedict's work and other published accounts, compares traits of 
personality of Japanese, Americans, and Europeans. No synthetic 
analysis approaching the stature of Benedict's work has, however, 
yet emerged. Conservative scholars, both Japanese and foreign, are 
well aware that the present state of knowledge of regional, class, and 
occupational differences makes generalizations on the Japanese per- 
sonality difficult, but the objective of an over-all characterization 
has not been cast aside. 

Content Analysis of Forms of Expressive Behavior 

Postwar publications by Japanese social psychologists have pre- 
sented a number of content analyses of Japanese movies, popular 
songs, life-counseling columns in newspapers, novels, and common 
folk-sayings, attempting to determine the values which stand out 
most strongly in these forms of expressive behavior.^ The technique 
is American derived, and in some instances the Japanese analysts 
have made comparisons with similar research in the United States. 
All of these studies have bearing on the subject of national charac- 
ter, although none attempts to be comprehensive in the manner of 
Benedict. We shall here present only a sample of the conclusions of 
these reports. 

The most ambitious of these impressionistic studies is Hiroshi 
Minami's Nihonjin no Shinri (Psychology of the Japanese) , which 
attempts to outline "those modes of feeling, thinking, and express- 
ing which are peculiar to the Japanese." Minami uses in a highly 
intuitive way popular songs, ideas expressed in fiction, common 
sayings, writings on army life, essays by successful men, and similar 
nonscholarly sources to deduce a number of themes or motifs. One 
wonders whether the themes are, in fact, inferred or whether the 
raw data are used to buttress preformulated themes. The work 
nevertheless contains observations that seem apt and, like others of 
its kind, provides information and interpretations that might serve 

^ Many of these have been translated into English. See Kato, ed. 1959. 


as starting points for future research. In a rather lengthy discussion 
of conceptions of happiness and unhappiness, for example, Minami 
observes that the Japanese seldom express happiness. Words con- 
veying this idea are few, and when they are used, the turn of expres- 
sion sounds awkward. The Japanese vocabulary is, however, rich 
in words denoting unhappiness. Many aphorisms, songs, writings, 
and personal philosophies of life contain as their central theme ways 
to cope with unhappiness, and attempts are made to justify unhap- 
piness on the grounds that it serves a useful purpose, as in ensuring 
the proper ordering of familial relations. Other major sections of 
Minami's work are entitled The Conception of the Self, Rationality 
and Irrationality, Spiritualism versus Sensualism, and Patterns of 
Human Relationships. No attempt is made to present a systematic 
characterization of the Japanese. 

An analysis of life-counseling columns in newspapers (Kato 
1959) reports that letters from the lovelorn are much fewer than 
is characteristic of similar columns in American papers. Letters are 
placed under three classifications: those concerned with group or 
international situations; those which center on human relations 
with one other individual ; and those expressing concern with height, 
weight, looks, and other physical features of the individual. Among 
adults the greatest source of distress is interpersonal relations in the 
family. Letters concerning relations between two individuals are 
principally between a young male and a young female. Among 
young girls the greatest concern is expressed over their own physi- 
cal features. A majority of letters from mature adults consist of 
complaints made against persons of higher social status than the 
writers. Wives complain more about husbands than husbands do 
about wives. This observation, it may be noted, seems contrary to 
the stereotype of the uncomplaining Japanese woman. 

Ananalysisof the lyrics of 61 postwar songs (Kato 1959), judged 
to be the most popular on the basis of sales of phonograph records, 
reports that the majority are sentimental, telling of love 
but never expressing happy sentimentality. The authors find in the 
songs four prevailing motifs: pessimism, fatalism, "existentialism" 
(explained as unexpressed feelings of loneliness and helplessness), 
and "premodern humanism" (feudal values in interpersonal rela- 

A useful review and critique of early postwar Japanese writings 
of similar kind that relate to traits of the Japanese personality has 
been made by Dore (1953)- 

JAPAN 2 5 

Studies Using Projective Techniques 

Among the techniques of personahty and culture research, Japa- 
nese scholars have made by far the greatest use of projective tests.'"* 
Unfortunately, analysis has generally been confined to interpreta- 
tion of the tests themselves with little or no attempt to relate find- 
ings to elements of culture. Results of most studies using the 
Rorschach, for example, are relatively crude statistics on the types 
of responses, giving means and percentages of color, movement, ani- 
mal content, and whole responses. During World War II and shortly 
afterward a number of studies were conducted by Japanese scholars 
with a Rorschach in which certain standard blots were modified and 
new ones added. It is, of course, highly doubtful that the results of 
these studies can be directly comparable with those based upon the 
standard Rorschach. For lack of other opportunity to learn, many 
Japanese researchers using the Rorschach and other projective tests 
in normative, nonclinical studies have been self-taught from reading 
American publications, and their interpretations often indicate a 
lack of familiarity with the potentials and limitations of the tech- 
niques. It must be added that these scholars were sometimes emulat- 
ing the manner of use of projective tests followed by a number of 
American anthropologists some years ago. 

Another weakness of Japanese scholars employing projective tests 
has been a general reluctance to interpret findings of the tests. Even 
when interpretation is made, the basis for the conclusions presented 
is seldom clearly stated. Thus, although a number of studies attempt 
to depict modal personalities for individual villages or occupational 
groups, and many others describe types of responses, they are gen- 
erally of little value except insofar as they might constitute accept- 
able raw data. 

Recent research using projective tests appears to be more promis- 
ing, and we have already noted two of the major projects which 
employ them. The Human Relations Research Group at Nagoya 
University is presently preparing a report of the Rorschach tests of 
over 700 urban and rural residents. This report is probably unique 

Formosan natives were given Rorschach tests by a Japanese psychologist in 1930. This is said 
to constitute the earliest trial of the Rorschach on a primitive people. A fairly extensive program 
of psychological testing of Formosan aborigines was conducted from that time until World War II. 
Projective tests have also recently been given to the Ainu by Japanese researchers. In very recent 
years research by Japanese social scientists has again expanded into areas outside Japan, and projective 
tests have been used on native populations of Nepal, Thailand, Brazil, Peru, and several other 
countries. Analyses of most of these data have not as yet been made or have not been published. 


in the field of culture and personality because it is based upon re- 
search using many projective and nonprojective techniques, and 
is the first large sample of its kind that crosscuts occupational 
groups and social classes of a culturally elaborate, highly stratified 
society. Findings with the Rorschach (T. Murakami and others, 
personal communications) indicate that regional and class differ- 
ences within the area tested, central and southwestern Japan, are 
generally slight. On other tests, however, certain significant dif- 
ferences appear — as noted in the pages that follow. Reports by Japa- 
nese scholars on smaller Rorschach samples from other regions of 
Japan support the interpretations of the Nagoya group. A recent 
report (Kodama 1953) listing popular responses to the Rorschach 
by Japanese adolescents of the Tokyo area, for example, is strikingly 
similar to the findings of the Nagoya University Human Relations 
Research Group. Japanese responses to the Rorschach indicate char- 
acteristics markedly different from those regarded as general for the 
population of the United States. They may be summarized briefly 
as follows: 

The number of responses is low in all social groups. Rejections are very high 
(from 20 to 25 per cent) on colored card 9, and black and white cards 6 and 7. 
There is a relatively high rate of rejection of card 10, which seems related to an 
inability or reluctance to use the details on this complex card. Difficulty in han- 
dling color freely and other indications attest to difficulty with spontaneous affect. 
Although markedly lower among urban residents than among rural residents, per- 
sonal rigidity is generally very high in comparison with norms for the United 
States. A great deal of organizational drive in the use of intellectual functions is 
indicated; the Japanese subjects are prone to push for complex, integrated whole 
responses. The sense of reality is generally very adequate. Although sometimes 
imaginative, responses include little fantasy of an extreme sort in directions con- 
sidered primitive or psychopathological. The form level is characteristically quite 
high. Labile color responses are usually perceptually tolerated when they are in- 
corporated in some complex overall concept. Pure color by itself is almost com- 
pletely lacking. These and other signs attest to the effectiveness of ego control 
that appears to be characteristic throughout the population. 

Although less commonly used than the Rorschach until recently, 
other projective tests have been employed to interpret Japanese 
values and attitudes as well as personality dynamics, and they have 
yielded interesting results. Basing his arguments principally on re- 
sponses to the Thematic Apperception Test and a problem-situation 
test, De Vos (1960a) argues against the widely held view that Japa- 
nese culture may best be regarded as a "shame" culture in a guilt- 
shame dichotomy. He holds that the strong achievement drive so 
often noted among the Japanese is not to be understood solely in 


terms of shame-oriented concern with community standards, but 
is also hnked with a deep undercurrent of guilt. The Japanese seem 
to suffer from guilt which is not associated with any complex of 
supernatural sanctions, but is instead derived from the system of 
loyalties which cements the structure of their traditional society. 
Guilt in Japanese is hidden from Western observation because we 
do not understand Japanese familial relationships, and because con- 
scious emphasis on external sanctions helps to disguise the under- 
lying feelings of guilt which, severely repressed, are not obvious to 
the Japanese themselves. The keystone toward understanding Japa- 
nese guilt is held to be the nature of interpersonal relationships 
within the Japanese family, particularly the relations of children 
with the mother. The Japanese mother, without conscious intent, 
has perfected techniques of inducing guilt in her children by such 
means as quiet suffering. She takes the burden of responsibility for 
their behavior and, as also with bad conduct on the part of her hus- 
band, will often manifest self-reproach if her children conduct 
themselves badly or in any way fail to meet the standards of success 
set for the community. If one fails to meet social expectations, he 
thereby hurts his mother, and he also hurts other familial members; 
as a result, he suffers unhappiness and feelings of guilt. ^ 

Another study based upon responses to the Thematic Appercep- 
tion Test (De Vos and Wagatsuma 1959) reports a high incidence 
of concern over death and illness, which the authors interpret as 
introjection of guilt. Death and illness of parents, as seen in cards 
of the Thematic Apperception Test by respondents, is very often 
related by them in stories to failure of a child to comply with paren- 
tal wishes in entering an arranged marriage, or in meeting other 
standards of behavior and achievement. Another recurrent theme 
found in responses is that of expiation; achievement of honor or 
success on the part of a child atones for egocentric or profligate 
behavior. The manner of introjecting guilt among the Japanese is 
thus seen to be related to the strong drive toward achievement that 
Western observers have long noted and pondered upon. The Japa- 
nese interpretation of the meaning of illness is also contrasted by 
the authors with that of various groups of American Indians who, 
in attributing illness to witchcraft, make use of the mechanism of 

An unpublished report on Japanese attitudes toward arranged 
marriages (Wagatsuma and De Vos) analyzes responses to the 

*For another approach to the subjects of shame and guilt, see Hsu (1949). 


Thematic Apperception Test and compares them with data derived 
by techniques ehciting more consciously controlled attitudes. Al- 
though current public opinion in Japan is increasingly lenient to- 
ward love marriage, as opposed to the traditional arranged marriage, 
individuals who have contracted love marriages are often reported 
to feel considerable guilt and inner restriction. Dependent upon the 
level of consciousness involved, attitudes and emotional reactions 
toward the two forms of marriage differ. A phenomenon labeled 
"psychological lag" appears to exist. In responses to the Thematic 
Apperception Test many respondents give clear evidence of strong 
internalized feelings against love marriage, although, as revealed 
by opinion surveys and direct interviews, when speaking on a con- 
scious level these individuals express approval of this form of union. 

The Thematic Apperception Test has also indicated differences 
in attitudes between occupational and social groups that conform 
with and amplify observations made by ethnologists using tradi- 
tional techniques of interviewing and observation. A farm com- 
munity, in which the so-called "traditional" Japanese pattern of 
hierarchical authority according to age, sex, order of birth, and 
status in the household is well established, is compared with a fishing 
community, where social relationships within the family do not 
follow such a strict hierarchy (De Vos & "Wagatsuma, in press). 
Responses to tests indicate markedly less rigidity, freer expression of 
aggression between the sexes, and less guilt in connection with intra- 
f amilial relations among people in the fishing community. 

Projective tests have also been put to use in the study of Japanese 
communities abroad, and, to a lesser extent, in research on child 
training. Discussion of these studies follows. 

Sfudies of Early Socializotion 

Japanese customs of rearing and socializing children have been 
the focus of more research in the field of personality-and-culture 
than any other subject. Perhaps the outstanding feature of published 
accounts resulting from this research has been conflict of opinion. 
The principal controversy in the entire field of Japanese culture- 
and-personality has revolved about interpretations of the influence 
of practices of child rearing on the adult personality. Early wartime 
studies conducted in the United States emphasized customs of toilet 
training and weaning, and contended that Japanese practices, par- 
ticularly in toilet training, were harsh and strongly influenced the 
adult personality. In this as well as other instances where interpre- 


tatlons have conflicted, differences by region and class, changes in 
practices, and cultural influences other than child training were 
overlooked or ignored. Haring's (1953) observation seems note- 
worthy here. The Japanese personality, he states, is what might be 
expected of the people of a police state. 

The pioneer studies of Gorer and LaBarre, long looked upon with 
question, were based upon information drawn from a limited num- 
ber of informants residing in the United States who appear to have 
held middle-class ideas of child training current at that time. The 
results of an investigation of practices of toilet training among 
Hawaiian Japanese (Sikkema 1948) presented conflicting data, and 
cast further doubt on the idea that severity of toilet training con- 
tributed to the compulsive personality traits of the Japanese. The 
sample in this instance was composed of individuals stemming prin- 
cipally from rural Japan, who had presumably been exposed to 
American ideas. 

More recently, Betty Lanham (1956) has reported on a fairly 
extensive investigation in a community of southwestern Japan on 
practices of weaning, toilet training, and forms of sanctions used 
to discipline children. Her statements, based upon a questionnaire 
devised and administered by Japanese associates, generally agree 
with unquantified observations made by Margaret and Edward 
Norbeck (1956) in a fishing community approximately 200 miles 
from Lanham's community. Miss Lanham concludes that although 
there are a number of sharp differences between Japan and the 
United States in other customs of child training, practices of toilet 
training differ little. 

Lanham's report has been criticized by Japanese scholars, who 
report different findings. The greatest point of dispute has been 
practices of weaning. Japanese scholars (e.g., Hoshino, Sofue, and 
others 1958) have expressed doubt about Lanham's information. 
Basing their statements upon field investigations of their own in 
Nagano Prefecture and, especially, upon huge samplings by pedia- 
tricians in the Tokyo area, they find that weaning begins and ends 
earlier than Lanham reports. Part of the argument here appears 
to hinge on the definition of weaning. Japanese scholars hold the 
view the weaning begins with the introdtLCtion of supplementary 
"solid" foods, and thus the span of time from the beginning of 
"weaning" until the child ceases to nurse is long. Research planned 
or presently under way by Lanham, Sofue, Hoshino, and others 
should do much to clear up points of contention in this and other 
matters of child training. 


One of the more noteworthy of the Japanese studies that does take 
cognizance of differences by social class (Ishiguro 1955) describes 
practices of child training in three Japanese social strata called "old 
middle class," "new middle class," and "lower class," and compares 
these practices with those reported for the United States. As Lan- 
ham also notes, the nursing period in Japan is reported to be longer 
than in the United States, and nursing tends to be on demand rather 
than on a fixed time schedule. Toilet training begins and ends earlier 
in Japan, but, unlike circumstances in the United States, control 
over urination precedes bowel control. In both countries weaning 
is abrupt in approximately 20 per cent of the cases reported. Prac- 
tices of the American lower class are reported to resemble most those 
of the Japanese "old middle class," and practices of the American 
middle class are most similar to those of the Japanese "new middle 
class" and "lower class." Although this study recognizes that change 
has occurred in customs of child training (the category "new mid- 
dle class" is composed of salaried men in industry, commerce, and 
public service, a relatively new social group in Japan) , it is depend- 
ent upon the recall of the mothers who served as informants, and is 
thus subject to distortion — probably in the direction of modern 
trends of change in these practices. 

A subject of recent investigations by Japanese scholars has been 
the psychological effect of the ejiko, a type of cradle for children 
used over a wide area of rural Japan. The most common type of 
ejiko is made of straw and is bowl-shaped. When it is necessary for 
the mother to leave the child unattended, it is placed in a squatting 
position within the cradle, wrapped in a quilt, and tied by a rope 
so that hardly any movement of the body is possible. Preliminary 
papers on the distribution and local varieties of the ejiko have been 
published ( Sof ue 1958; Sue 1958; Sof ue. Sue, and Murakami 1958). 
An intensive study directed toward determining its psychological 
significance was conducted in 1958 and 1959 by Sofue and others 
in a hamlet of Nagano Prefecture. It is interesting to note that prac- 
tices of child training differ with social class even in this small rural 
community (Sue, personal communication) . This project includes 
the use of projective tests, and the data gathered will be compared 
with those obtained from other communities. It is not clear how the 
researchers intend to relate the findings of the tests to the custom 
of using the ejiko, or how the possible effects of use of the ejiko may 
be distinguished from those of other childhood experiences. 

Increasingly, both Japanese and American scholars engaged in 
research on Japanese practices of child rearing, as related to the 

JAPAN 3 1 

adult personality, have come to realize the weaknesses of an ap- 
proach that deals with formal customs such as toilet training or the 
use of the cradle. They have looked to multiple and less formalized 
factors, including the identity of the adults concerned in the sociali- 
zation process and affective relations between the socializers and 
the socialized. Greater attention is now given to such questions as 
the length of time the child sleeps with its parents, who bathes a 
child or accompanies it in the bath, and the manner of gratifying 
impulses (e.g., Caudill 1959c). 

An indication of the multiplicity of factors involved in the for- 
mation of the adult personality is provided by the results of a psy- 
chological testing of Kihei, American-born Japanese who, after 
spending their early childhood in the United States, are taken to 
Japan for a number of years for schooling, and then return to the 
United States (De Vos 1955) . From the standpoints of personality 
rigidity and maladjustment, the Kibei were generally intermediate 
to the Issei and Nisei. If the earliest practices of socialization are in 
fact the most powerful, little difference should of course be found 
between Kibei and Nisei, as they appear to have been exposed to 
essentially identical practices of training in infancy and early child- 
hood. Assimilation of Japanese values later in childhood and during 
adolescence seems to be the source of conflict for the Kibei. 

Although not focused directly on the subject of customs of child 
rearing, research presently being conducted by Ezra Vogel on the 
linkage between intraf amilial social relations and emotional disturb- 
ances has much relevance. The project consists of the intensive study 
of familial relations among the members of twelve Japanese families 
of comparable social and economic backgrounds, of which six have 
"normal" children and the remaining six have one or more emo- 
tionally disturbed children under intensive treatment at the Japa- 
nese National Institute for Mental Health. Vogel reports that the 
emotional attachment of the Japanese child to his family is greater 
than that of the American child and is maintained for a longer pe- 
riod, and that tensions arising out of relationships with kin are more 
common in Japan. Conflicts within the family follow a limited 
number of characteristic patterns, such as tension between a man's 
wife and his mother, and tension on the part of the wife because of 
the husband's habit of seeking sexual and other gratifications out- 
side the home. The degree and type of conflict among adults are 
related to the intensity and type of emotional disturbance of the 
children. This research is organized so as to allow direct comparison 


with similar studies of the Harvard Psychological Clinic on fa- 
milial relations among Irish-Americans, Italian-Americans, and 
old Americans. 

Somewhat more peripheral to the subject of child rearing is Wil- 
liam Caudill's current research on the subject of impulse gratifica- 
tion and restraint. Basing his statements on responses to a picture 
test similar to the Thematic Apperception Test, Caudill (1959c) 
reports that there are differences between Japanese and Americans 
in what is ego-syntonic (consciously acceptable to the ego and need- 
ing no repression) . For example, a Japanese mother's sensual grati- 
fication in nursing her infant is consciously acceptable to her, 
whereas the feeling of gratification is generally repressed by the 
American mother. The Japanese are also described as being more 
ego-syntonic with reference to certain forms of mutual dependency 
within the family. A young man, for example, may remain depend- 
ent upon his mother for many satisfactions long past the age that 
would be considered appropriate in the United States. The Japanese 
are said to be much less ego-syntonic than Americans in direct ex- 
pression of aggression. Caudill (1959b) relates the hypochondriasis 
manifest in the Japanese to their inability to express direct aggres- 
sion toward others easily and the consequent deflection toward the 
self in various forms including hypochondriasis. 

An interesting and useful film on child rearing gives a visual com- 
parison of Japanese, Indian Hindu, French, and Canadian prac- 
tices (National Film Board of Canada 1959) • The Japanese section, 
prepared with the advice of William Caudill, depicts the events of 
a day in the life of an infant girl, 10 months of age, from a farming 
family of the Kanto Plain, near Tokyo. 

Other current research using a personality and culture approach 
and bearing upon the subject of child training is De Vos's study of 
juvenile delinquents in the Tokyo area (1960b) . Data were gath- 
ered by means of conventional interviews, Q-sort cards, the Thema- 
tic Apperception Test, and the Rorschach on the attitudes of juve- 
nile delinquents toward their parents and their conceptions of the 
social roles of mother and father. These data will be compared with 
findings on a control group of nondelinquent Japanese youth and 
with similar material previously gathered on delinquent and non- 
delinquent groups of juveniles, including Negroes and Mexican- 
Americans, in the United States. 

Japanese child psychologists and pediatricians have carried out a 
considerable number of studies of child training, but these are often 

JAPAN 3 3 

fragmentary, and they have not attempted to analyze systematically 
the interrelationships between child rearing and the formation of 
personahty. Little consideration is given in these studies to the social 
class of the informants used. The studies are also characterized by 
some degree of "culture-blindness." Much that is pertinent to an 
understanding of the relationships between personality and culture 
is overlooked simply because it is so familiar to the scholars them- 
selves that it escapes notice or is deemed unworthy of it. 

Although they have wider significance, a group of unique papers 
by the Japanese psychologist Takeo Doi (1956, 1958, i960) touch 
indirectly upon child training. Doi calls attention to Japanese words 
and concepts as illustrative of Japanese psychology, and states that 
terms referring to the emotions and interpersonal relationships 
often have no suitable English equivalents. He cites as an example 
the noun amae, derived from the verb amaertt, which he defines in 
English (1958, writer's English abstract) as fo depend and presume 
upon another's love or indulge in another's kindness. (A popular 
Japanese-English dictionary [Masuda 1957] translates this word, 
dependent upon context, as: to baby; to act like a spoiled child; 
to coax, to be coquettish; to faivn upon; and to avail oneself of an- 
other's kindness.) In Western psychological terms, Doi holds, the 
word amae (or aniaeru) has a central meaning referring to de- 
pendancy needs. To Japanese minds it usually means what a child 
feels about or how he acts toward his parents, particularly his 
mother, and thus it distinctly relates to the nursing period. Think- 
ing in terms of this familiar Japanese concept, Doi states, easily led 
Japanese psychoanalysis to formulate theories about the importance 
of oral dependency in the formation of neuroses, an interpretation 
which has only recently become the focus of psychoanalytic work- 
ers in Western nations. 

Studies of Japanese Overseas 

A group of studies which give promise of being particularly use- 
ful in a number of respects is that conducted on Japanese immi- 
grants to foreign countries and their descendants. Sociologists, 
educational psychologists, anthropologists, and scholars in other 
disciplines of social science have engaged in research of this kind, 
which has been concerned principally with Japanese in the con- 
tinental United States and Hawaii. Interest has grown to include 
Japanese in South America and Canada, where field research has 
recently been conducted or is now in progress, 


One of the earliest studies was an investigation by educational 
psychologists of school behavior of Nisei children on the Pacific 
coast. An extensive series of studies, most of them summarized by 
E. K. Strong ( 1934) , compares Caucasian-American and Japanese- 
American grade school and high school children in intellectual func- 
tioning and related features of personality. These studies fail to 
make any explicit use of the concept of culture and for this reason 
appear naive in the light of present-day theory and knowledge in 
the social sciences. They bring out distinctly, however, a number of 
traits that characterize the Nisei. Psychological tests indicated no 
differences in the intellectual functioning of Nisei and Caucasian- 
American children, although they did indicate different artistic 
sensibilities. The sense of composition and the use of line of the Nisei 
was found to be superior; their use of perspective inferior. Other 
traits noted are of greater interest. One of these is close conform- 
ance with middle-class American norms of behavior. Behavior of 
the Nisei children in the schools is described as characteristically 
docile, patient, and respectful and obedient to the teachers. Motiva- 
tion to achievement is strong, and it is clear that parents of the Nisei 
exerted strong pressure to inculcate in their children the idea that 
meeting American standards of achievement and other norms of 
behavior is desirable. Nisei students tended to earn higher grades 
for school work than others and to receive greater recognition from 
teachers for exemplary conduct. 

Interest in the Japanese of the Pacific coast and elsewhere in the 
United States was heightened during "World War II, when many 
were sent to relocation centers. A number of published accounts 
deal with the adjustment of the Japanese to life in these camps and 
the new surroundings to which they moved when the war ended 
and the camps were closed (e.g., Leighton 1945) , but they do not 
relate directly to the Japanese personality. 

Postwar interdisciplinary research on the acculturation of Japa- 
nese in the Chicago area has yielded publications on acculturative 
changes in personality and on the nature of psychological conflicts 
which the bridging of Japanese and American cultures has produced 
among the Japanese- Americans. Among these is Caudill's (1952) 
extensive analysis of psychological aspects of the drive toward 
achievement and other value attitudes of the Nisei. A study, based 
on Rorschach tests, of acculturative changes in structural aspects 
of personality of Issei and Nisei (De Vos 1955) reports a high level 
of rigidity and certain indications of maladjustment among the 

JAPAN 3 5 

Issei. Nisei were much lower in rigidity, and displayed fewer indi- 
cations of maladjustment. Comparison with data on Japanese in 
Japan of the same social backgrounds (i.e., rural residents) revealed 
equally high rigidity, but indications of severe maladjustment were 
found only among the American Issei and appear to be related to 
stress in adjusting to the alien American culture. 

A focus of continuing interest in the study of the Japanese in 
America has been attempts to analyze their drive toward achieve- 
ment. The question has been asked why Issei and Nisei have ap- 
parently adopted the attitudes and values, including the strong 
motivation to achievement, of the American middle class when cer- 
tain other immigrant groups under comparable circumstances have 
not done so to the same degree (see, for example, Norbeck 1959 
on ethnic groups in Hawaii) . Scholars have also asked why the Japa- 
nese have made such apparently successful adjustments to life in the 
United States when other minority groups, some of them suffering 
less social discrimination, have failed to do so. 

Similarities and compatibilities in certain American and native 
Japanese values and attitudes have been offered in partial explana- 
tion (Caudill and De Vos 1956). Japanese are described as ex- 
tremely sensitive to stimuli from the outer world and as having a 
superego structure that depends strongly on external sanctions for 
reinforcement. Cultural values are internalized in a socialization 
process that emphasizes long-range goals, perseverance, obedience 
to authority, and a sense of obligation to parents. Socialization takes 
place within the family, but the drive to achievement is satisfied by 
conforming with expectations of the outer society. (This observa- 
tion, it may be noted, is in keeping with opinions expressed by 
numerous other scholars. For example, a study comparing the 
vocational aspirations of American and Japanese schoolchildren 
[Goodman 1957] describes the Americans as "self-oriented" [ego- 
centric] and the Japanese as "others-oriented.") Attitudes and 
community values to which the Nisei, as a minority group, are most 
strongly exposed in extrafamilial contacts, and which the Nisei 
internalize, are those of the American middle class. Thus native 
Japanese and American attitudes of valuing conformance and 
achievement and stressing long-range goals reinforce each other. 
Success for the Nisei differs from success for the non- Japanese 
American, however, in being closely related to the fulfillment of 
filial obligations. The feeling of necessity to succeed as a means of 
satisfying obligations to parents is brought out in clinical studies 


of individual Nisei (e.g., Babcock and Caudill in G. Seward, eJ. 
1958) . A tendency toward psychological depression among Nisei is 
well documented in a collection of papers on culture conflict related 
to psychiatric problems of the Nisei (G. Seward, ed. 1958) , which 
includes a particularly pertinent paper by Marvin Opler on psy- 
chological stress as related to filial obligations in the case history of 
an individual Kibei. 

A study which compares acculturating Arabs in Algeria and 
other minority groups with Japanese- Americans reports that cer- 
tain indications of intrapsychic stress appear in the Rorschach rec- 
ord of all groups, although they are less marked among the Nisei 
than among the Jssei. Stress is seen to be connected with accultura- 
tion or status as members of minority groups because the indications 
do not appear in the records of individuals when they are members 
of a majority group (De Vos, in Kaplan, ed. 1961 ) . 

Data on immigrant and South American-born Japanese in Peru 
and Brazil that will allow comparison with studies in the United 
States have recently been collected under the direction of Seiichi 
Izumi of Tokyo University. Results of Rorschach, Thematic Ap- 
perception Tests, and problem situations tests are now in the process 
of analysis, and promise to allow direct comparison of personality 
traits and problems of acculturation among Japanese of the United 
States and these two South American countries. Preliminary analy- 
sis (Hiroshi Wagatsuma, personal communication) indicates that 
results of projective tests administered in Peru differ from those 
obtained in Brazil and the United States as well as Japan. Japanese- 
Peruvians appear to be less strongly motivated toward personal 
achievement than Japanese-Brazilians, Japanese-Americans, or 
Japanese in Japan, and, as indicated by the Rorschach, to be more 
pragmatic, presenting less emphasis on the integrated conceptions 
characteristic of the Japanese in the United States and at home. 

An interdisciplinary study under the direction of R. P. Dore of 
the University of British Columbia and Masao Gamo of Meiji Uni- 
versity on a fishing community of British Columbia populated by 
Japanese immigrants and their descendants is worthy of note. The 
aims of this study include investigation of problems of accultura- 
tion and research on personality, and the project includes projective 
tests among its tools. 

Although the R^^ukyu Islands are hardly "overseas" in the same 
sense as North and South America, research in personality and cul- 
ture on the inhabitants of these islands is of significance for com- 


parison in quite the same way as data on Japanese in faraway lands. 
Although to some degree culturally and perhaps physically distinct 
from the residents of Japan proper, the Ryukyu Islanders speak a 
Japanese dialect, regard themselves as Japanese, and are so regarded 
by the people of the principal islands of Japan. The Ryukyus were 
a part of the Japanese nation for many centuries before the end of 
World War II, but because of their isolated geographic position 
the islands escaped or were only lightly affected by many cultural 
innovations that swept Japan proper. In a provocative short article 
on the island of Amami Oshima, Douglas Haring (1954) describes 
the islanders as having more "open" personalities than the residents 
of Japan proper. He suggests that the lack of sustained direct con- 
tact with Japan proper prevented the spread to this small island of 
attitudes and values which permeated the principal islands during 
the Tokugawa era (1603 — 1868). The modern Amami Oshima 
islanders, more impulsively labile and directly expressive of emo- 
tions than modern mainland Japanese, may represent a type of per- 
sonality that characterized the whole nation before Tokugawa 

J. Moloney's controversial writings on the Okinawans describe 
them as relatively free of conflict and assert that as a result of per- 
missive practices of nursing there is little mental illness among them. 
A detailed field study of child-rearing practices in an Okinawan 
community conducted in 1957 by T. W. and H. S. Maretzki (per- 
sonal communication) casts much doubt on the statements of 
Moloney. Research by the Maretzkis centered on dependence- 
independence, aggression, and internalization of values with the 
objective of relating measures of children's personality to ante- 
cedent factors of socialization. They observe that both adults and 
children indulge in a great deal of verbal aggression, and they report 
many traits that differ from observations made on Japanese in the 
main islands. Notable among these is less parental stress upon 
achievement by their children. Tightly knit social relationships 
throughout the whole community, encouraged by the customs of 
community endogamy, are tied in with the high sociability and little 
concern with competitiveness which characterize the children. Out- 
standing features in Okinawan socialization include an emphasis 
on nurturance, a high diffusion of caretakers of children, and the 
importance of the role of peers in every stage of child development. 
The community as a whole is almost an extension of the household 
environment. Additional research by the Maretzkis in i960 is fo- 


cussed on the adult personality and will provide data for more de- 
tailed comparison with Japan proper as well as other countries. 

A study of another Ryukyu community on Ishigaki Island by 
Allan H. and Ann Gertrude Smith now in press (American Philo- 
sophical Society) provides additional data on Ryukyuan practices 
of child rearing, including mechanisms of social control. 

Pertinent investigations that fall outside the classifications we 
have used here, but are worthy of notice, include research on pat- 
terns of suicide in Japan and types of therapy used in Tokyo hos- 
pitals to treat psychopathology. Seiichi Kato of the National 
Institute for Mental Health (of Japan) has been conducting re- 
search for ten years on Japanese suicides and attempted suicides with 
the objective of determining their patterns and social correlates. 
It is interesting to note that statistics since 1882 show that incidence 
by age groups has been essentially constant until the end of World 
War 11. Postwar statistics reveal a rise in suicides among young 
males and, although figures for this group are still high, a decrease 
among young women. Caudill has completed field research (1959) 
on the psychiatric techniques and social environment of three To- 
kyo mental hospitals, one of which emphasizes organic therapy, 
another psychoanalytic treatment, and the third a distinctively 
Japanese form of treatment named Morita therapy after its founder 
and derived in part from Zen Buddhism. 

Summary and Conclusions 

An over-all view of research in Japanese culture and personality 
reveals both strengths and weaknesses. The absolute number of for- 
eign and native scholars engaged in research on Japan in this field is 
not great, yet few if any foreign cultures have been the subject of 
study by so many individuals. The total of published studies, many 
of them in the Japanese language, is impressive in volume, but it is 
weak or deficient in a number of respects, some of which we have 
noted. The techniques and theories of modern culture and per- 
sonality research have as yet hardly had adequate testing in Japan — 
but it is extremely doubtful that they have had adequate testing in 
any other culture. Considerable progress has been made in deter- 
mining differences in culture and traits of personality according to 
region and social class in Japan, but much more is required before 
generalizations on the nation may be made with assurance. A con- 
spicuous failing of Japanese scholars has been concentration on 
minute problems and a reluctance to go beyond mere description. 


It must be added that research by Japanese scholars has been greatly 
inhibited by lack of funds, and their emphasis on studies of small 
scope is in part due to this circumstance. Japanese scholars have also 
been at a serious disadvantage for lack of opportunity to receive 
training in the techniques of personality and culture research. Only 
in very recent years have a few had the opportunity to take profes- 
sional training. 

Despite these negative comments, research in Japanese culture- 
and-personality has not been merely a spotty repetition of tech- 
niques and interpretations borrowed from scholars of the United 
States and Europe. It has made its own contributions of theoretical 
significance and it holds unusually great promise of making future 
contributions. Subjects of research have been examined in such a 
way that their conclusions concern and shed light on issues of gen- 
eral interest in the field of culture-and-personality and the social 
sciences as a whole. Past or present research in Japanese culture-and- 
personality has special relevance to the following subjects of general 
interest : 

1. The nature of human drives to achievement. 

2. Variations in the cultural conditioning of basic psychological mechanisms: 
shame versus guilt as motivating forces; different uses of introjection and pro- 

3. Processes of acculturation: factors involved in making acculturation easy 
and successful or difficult and unsuccessful; the relationship between accultura- 
tion and psychic stress. 

Motivation toward achievement has long been a subject of schol- 
arly interest, and the practical value of an understanding of factors 
that inhibit and encourage the growth of drives to achievement is 
obvious. Explanations have been sought through examination of 
religiously sanctioned ideals of behavior and in many other ways. 
The eagerness and speed with which Japan assimilated Western 
culture, the startling rapidity with which it rose to a position as a 
major international power, and the remarkable recovery of the na- 
tion after devastating defeat and economic collapse in World War 
II have stimulated much curiosity and theorizing. Historians have 
pointed to the long-established receptiveness of the Japanese to 
items of foreign culture and their equally long record of successful 
adaptation of borrowed items. Other scholars have held that the 
hierarchical ordering of Japanese society, especially the former 
tight control of ruler over subject, has made the industrialization 
and "modernization" of Japan easy. Robert Bellah's recent and in- 


teresting Tokugawa Religion (1957) approaches the problem so- 
ciologically after the manner of Max Weber. He concludes that an 
equivalent of the Protestant ethic, evident in Tokugawa times, 
served as a spur to Japanese economic growth. T. C. Smith (1959) 
has argued effectively that the road to industrialization was paved 
by indigenous developments during Tokugawa times. 

All of these studies leave off where culture-and-personality be- 
gins. The pattern of psychological integration of the personality 
that encourages diligence and self-denial for the purpose of attain- 
ing long-range goals is of particular interest in understanding the 
achievements of Japan as a nation. It is here that research in culture- 
and-personality can be very helpful. As we have noted, much evi- 
dence from studies in culture-and-personality indicates that strong 
motivation toward success exists among the Japanese of Japan and 
Japanese-Americans. Other research has suggested the means by 
which motivation is inculcated and reinforced. Research under way 
on intrafamiiial relations gives promise of telling us more about 
motivation as it is related to Japanese social structure as well as 
contributing to our understanding of psychological stress arising 
from social living. These theoretical matters are, of course, highly 
relevant to the problem of understanding other Asian countries 
where economic developments have followed quite different courses, 
and to the understanding of motivation and achievement for all 

In connection with the problem of understanding the drive to 
achievement of the Japanese, published studies in personality and 
culture have presented hypotheses that should stimulate re- 
examination of theories of the relationships between superego and 
ego ideal as these are related to guilt and shame. Perhaps all scholars 
working in the field of culture-and-personality would agree that 
theorizing on the subject of guilt versus shame has often been over- 
simplified. Certainly, the Japanese studies suggest strongly that 
shame and guilt are not necessarily antithetical or mutually incom- 
patible. The question of the weighting of the sanction of shame 
versus that of guilt in any society cannot be investigated satisfac- 
torily without consideration of several other related subjects, in- 
cluding the mechanisms of introjection and projection. Research 
on Japan on this latter subject points up the necessity of re- 
examination of theories and of further cross-cultural comparison. 

Perhaps the most promising avenue of research in Japanese per- 
sonality and culture bears on the subject of acculturation. The fact 


that Japanese citizens of similar backgrounds have migrated to sev- 
eral nations with quite different cultures provides a unique oppor- 
tunity for cross-cultural comparison of processes of acculturation. 
Research completed to date indicates that the Japanese of the United 
States differ considerably in traits of personality from those who 
have settled in South America. Studies of the Japanese in these areas 
suggest that compatibility rather than duplication of values be- 
tween the minority and majority group are necessary for successful 
acculturation, and that quite different patterns of psychological re- 
inforcement of values may yield results that are similar. Delineation 
of the values as well as interpretation of associated psychological 
mechanisms are problems which appear to yield best results when 
approached through the methods of personality and culture. 

Research conducted to date on the Japanese also indicates that 
projective tests are useful instruments for detecting intrapsychic 
stress arising from difficulties of acculturation. Further comparison 
with data on Chinese-Americans, American Negroes, Puerto Ri- 
cans, Filipinos, and other minority groups and acculturating peoples 
in the United States in this and other matters should be extremely 

The promise which future research holds seems particularly great. 
Japan is a large and culturally complex society with many social 
strata representing subcultures, and many regional differences. This 
circumstance provides an unusually fine opportunity for compari- 
son to aid in gaining understanding of many questions concerned 
with personality and culture. The Japanese abroad offer another 
useful avenue of comparison. Japan is, moreover, a highly literate 
society with much recorded history. During the past century it has 
undergone tremendous cultural change, proceeding at an acceler- 
ated rate since the end of World War II, and much of the change 
is well documented. In these respects, Japan offers an exceptional 
opportunity for observation of sociocultural change and its rela- 
tionship to personality. In all of these matters, the prospect of fu- 
ture contributions to knowledge is particularly favored by the fact 
that both native and foreign scholars in several disciplines are en- 
gaged in research directed toward solving the same problems. 


(Titles of Japanese works have been Romanized, and English 
translations are given. When Japanese journals and serials have offi- 


cial English as well as Japanese titles, both are cited. English abstracts 
are published with some Japanese articles, and these have been 

Babcock, Charlotte, and William Caudill 

1958 Personal and cultural factors in treating a nisei man. In Clinical studies 
in culture conflict, Georgene Seward, ed. New York, Ronald Press. 
Bellah, Robert N. 

1957 Tokugawa religion; the values of pre-industrial Japan. Glencoe, Free 
Benedict, Ruth 

1946 The chrysanthemum and the sword: patterns of Japanese culture. Bos- 
ton, Houghton Mifflin. 
Bennett, J. W. and Michio Nagai 

1953 Echoes: reactions to American anthropology — Japanese critique of the 
methodology of Benedict's "Chrysanthemum and the sword." Ameri- 
can Anthropologist 55:404-411. 

Buchanan, D. C. 

1954 Japanese character and personality as revealed in their culture, ht Un- 
derstanding other cultures, William A. Parker, ed. Washington, Ameri- 
can Council of Learned Societies. 

Caudill, William 

1952 Japanese-American personality and acculturation. Genetic Psychology 

Monographs 45. Provincetown, Mass., Journal Press. 
1959a The relationship of anthropology to psychiatry in the study of culture 
and personality. Seishin Bunseki Kenkyu (The Japanese Journal of 
Psychoanalysis) 6:57—65. 
1959b Similarities and differences in psychiatric illness and its treatment in 
the United States and Japan. Nagoya University, Seishin Eisei (Mental 
Hygiene) 61/62:15-26. 
1959c Watakushi no piknchd intabyil gipitsu (The use of a "Picture Inter- 
view" technique in the study of impulse gratification and restraint). 
Yokohama, Hiyoshi Byoin, Seishinbunsekigaku no Susume 3:1—13. 
Caudill, William and George De Vos 

1956 Achievement, culture and personality: the case of Japanese Americans. 
American Anthropologist 58:1102-1126. 
CoLTON, H. A., Jr. and F. G. Ebaugh 

1946 Japanese neuropsychiatry. American Journal of Psychiatry 103:342- 
Dening, Walter 

1 891 Mental characteristics of the Japanese people. Trans, and Proc. of the 
Japan Society, old series, 19(1) : 1 7-3 6. 
De Vos, George 

1954 A comparison of the personality differences in two generations of Japa- 
nese Americans by means of the Rorschach test. Nagoya Journal of 
Medical Science 17 (3 ) :i53— 265. 

1955 A quantitative Rorschach assessment of maladjiostment and rigidity In 


acculturating Japanese Americans. Genetic Psychology Monographs 
52 (First Half) : 5 1-87. Provincetown, Mass., Journal Press. 

1960a The relation of guilt toward parents to achievement and arranged mar- 
riage among the Japanese. Psychiatry: Journal for the Study of Inter- 
personal Processes Vol. 23, No. 3. 

1960b Psycho-cultural attitudes toward primary relationships in Japanese de- 
linquents — a study in progress. Seishin Eisci (Mental Hygiene) , No. 66. 

1961 Symbolic analysis in the cross-cultural study of personaUty. In Studying 
personality cross-culturally, Bert Kaplan, ed., Evanston and White 
Plains, Row Peterson. 
Df. Vos, George and H. Wagatsuma 

1959 Psychocviltural significance of concern over death and illness among 
rural Japanese. International Journal of Social Psychiatry 5:5-19. 

1 96 1 Variations in traditional value attitudes toward status and role behavior 
of women in two Japanese villages. American Anthropologist, in press. 
Doi, Takeo 

1956 Japanese language as an expression of Japanese psychology. Western 
Speech 20:90—96. 

1958 Shinkeishitsti no sehhinbyori (Psychopathology of "shinkeishitsu") . 
Seishinshinkeigaku Zasshi (Psychiatria et Neurologia Japonica) 60:733— 
744 (English abstract). 

i960 Jibtm to amae no seishinbyori (Psychopathology of "jibtm" and 
"amae"). Seishinshinkeigaku Zasshi (Psychiatria et Neurologia Ja- 
ponica) 62:149—162 (English abstract) . 
Embree, J. F. 

1939 Suye Mura, a Japanese village. Chicago, University of Chicago Press. 


1952 Rorschach test ni yoru personality no chosa (I) — Nara-ken, Seiki-gtm, 
Hirano-mtira no baai (An investigation of personality by means of the 
Rorschach test, I, Hirano Village, Seiki-gun, Nara Prefecture). Ky5to 
Daigaku Jimbun Kagaku Kenkyusho Chosa Hokoku (Social Survey 
Report of The Research Institute for Humanistic Studies, Kyoto Uni- 
versity) , 8. 

1957 A statistical approach to group comparison based on the distribution 
of Rorschach responses. Memoire of The Research Institute for Hu- 
manistic Studies, Kyoto University. 

1958 Jinruigaku ni okerti personality no mondai — Rorschach test ni yorn 
hikaku kenkyii (Some problems of personality studies in anthropology 
— comparative research by means of the Rorschach test). Shiso 

1959a Rorschach hanno no ingakuteki bumpu ni yoru group hikaku no koko- 
romi (An attempt to compare groups by means of the mathematical dis- 
tribution of Rorschach responses). Shinrigaku Hyoron (Psychological 
Review) 1:35—49. 

1959b Rorschach hannoshn — Nihon nosanson dansei shotaishu no baai (Tables 
of Rorschach responses of male house-holders in Japanese farming and 
mountain communities). Kyoto Daigaku Jimbun Kagaku Kenkyusho 
Chosa Hokoku (Social Survey Report of the Research Institute for 
Humanistic Studies, Kyoto University), i8. 


FujioKA, Y., Y. Maki, T. Ikeda, and M. Okano 

1956 Rorschach test ni yoru personality no chosa (III) — Nara-ken, Yoshino- 
gun, Totsngawa-viura no baai (An investigation of personahty by means 
of the Rorschach test, III, Totsugawa Village, Yoshino-gun, Nara Pre- 
fecture) . Kyoto Daigaku Jimbun Kagaku Kenkyusho Chosa Hokoku 
(Social Survey Report of The Research Institute for Humanistic Studies, 
Kyoto University) , 14. 

Goodman, M. E. 

1957 Values, attitudes and social concepts of Japanese and American children. 
American Anthropologist 59:979—999. 

GoRER, Geoffrey 

1942 Japanese character structure and propaganda: a preliminary survey. 
Prepared for the Committee on National Morale and the Council on 
Human Relations. Yale University (mimeo). 

1943 Themes in Japanese cultvire. Trans. New York Academy of Sciences, 
Series II, 5:106—124. 

Haring, D. G. 

1943 Comment on Japanese personal character. Excerpts from Blood on the 

rising sun, by D. G. Haring. 

Reprinted in Personal character and cultural milieu, D. G. Haring, ed. 

Syracuse, Syracuse University Press. 
1946 Aspects of personal character in Japan. Far Eastern Quarterly 6:12—22. 
1949 Japan and the Japanese. In Most of the world, Ralph Linton, ed. New 

York, Columbia University Press. 
^953 Japanese national character; cultural anthropology, psychoanalysis and 

history. The Yale Review 42:375-402. 
1954a Comment on field techniques in ethnography; illustrated by a survey 

of Amami Oshima. Trans. New York Academy of Sciences 16:271- 

1954b Comment on field techniques in ethnography; illustrated by a survey 

of the Ryukyu Islands. Southwestern Journal of Anthropology 10:255- 

HosHiNO, Akira, Takao Sofue, Hiroko Sue, and Yoshikazu Imai 

1958 Ikii]i yoshiki to paasonaritei (Infant training and personality) (I), In- 
ternational Christian University, Kydiku Kenkyu 5:148-216. 

Hsiao, H. H. 

1939 Mentality of the Chinese and Japanese. Journal of Applied Psychology 

Hsu, F. L. K. 

1949 Suppression versus repression: a limited psychological interpretation of 
four cultures. Psychiatry 12:223-242. 

Imanishi, K. 

1952 Mura to ningen (Villages and people). Tokyo, Shinhy5ronsha. 

Imanishi, K., Y. Maki, and Y. Fujioka 

1955 Rorschach test ni yoru personality chosa (A study of personality by 
means of the Rorschach test). In Taclfikuiyo no kenkyu — gijiitsJi, 
seikatsu, ningen (Studies on a pottery-making village, Tachikui — tech- 
nology, way of life, the people), K. Yabuuchi, ed. Tokyo, Koseisha 



1955 Haha-ko kankei no shinrigakicteki kenkyu {sonoichi) — Nyuyoji-kJ no 
shitsukekata no jittai (Psychological study on mother-child relations 
(I) — actual circumstances in child training). Nagoya Daigaku Kyoiku 
Gakubu Kiyo (Bulletin of the Faculty of Education, Nagoya Univer- 
sity) 1:74-86 (English abstract). 

Kaplan, Bert, ed. 

196 1 Studying personahty cross-culturally. Evanston and White Plains, Row 
Kato, Seiichi 

1956 Suicide. In Annual Report of Mental Health of the National Institute 
of Mental Health, Chapter IV, Section 22. Tokyo. 

Kato, Hidetoshi, ed. and trans. 

1959 Japanese popular culture. Tokyo and Rutland, Vt., Charles E. Tuttle. 
Kawashima, Takeyoshi 

1951a Giri no kannen ni tsuite (On the concept of giri) . Sbiso, Sept., 21-28. 

1951b On no is/yiki no pttai (The nature of the concept of on). Chuo Kdron 

1957 Ideorogii to shite no kazokti seido (The family system as ideology). 
Tokyo, Iwanami Shoten. 

Kerlinger, F. N. 

1953 Behavior and personality in Japan: a critique of three studies of Japa- 
nese personality. Social Forces 31:250—258. 


1956 Nihonbiinka no kontei ni hisoviiimono (What lies at the bottom of 
Japanese culture). Tokyo, Kodansha. 


1953 Nihonpn no ryorusbyakku banno no kenkyii (A study of Rorschach 
responses of Japanese). Shinrigaku Kdza 7:1-92. 

Kyugakkai Reng5 Tsushima Kyodo Chosa Iinkai 

1954 Tsushima no shizen to bunka (Nature and culture in Tsushima). 
S5g5 Kenkyu Hokoku, 2. Tokyo, Kokinshoin. 

LaBarre, Weston 

1945 Some observations on character structure in the Orient: the Japanese. 
Psychiatry 8:319-342. 
Lanham, Betty B. 

1956 Aspects of child care in Japan: preUminary report. In Personal char- 
acter and cultural milieu, D. G. Haring, ed. Syracuse, Syracuse Uni- 
versity Press. 

Leighton, Alexander H. 

1945 The governing of men; general principles and recommendations based 
on experience at a Japanese relocation camp. Princeton, Princeton Uni- 
versity Press. 

Maretzki, T. W. 

1957 Child rearing in an Okinaw^an community. Yale University, Ph.D. dis- 

Masuda, K. 

1957 Kenkyusha's new pocket Japanese-English dictionary. Tokyo, Kcnkyu- 


Meadow, Arnold 

1944 An analysis of Japanese character structure. Prepared for and dis- 
tributed by the Institute for Intercultural Studies, New York City 


1954 Ni/jonjin no sbinri (Psychology of the Japanese). Tokyo, Mainlchi 
Moloney, J. C. 

1945 Psychiatric observations on Okinawa Shima. Psychiatry 8:391-399. 
195 1 A study of neurotic conformity: the Japanese. Complex 5:26-32. 

1953 Understanding the paradox of Japanese psychoanalysis. International 
Journal of Psychoanalysis 34 (part 4) :i-i3. 

1954 Understanding the Japanese mind. New York, Philosophical Library. 
Mori, Shigetoshi and Tadashi Miwa 

1958 Okmoerabiiid-iomin no paasonaritei (Personality of Okinoerabuto 
Islanders). Jinruikagaku, 10. 

National Film Board of Canada 

1959 Four families (film on practices of child rearing). Box 6100, Montreal, 

NoRBECK, Edward 

1959 Pineapple town — Hawaii. Berkeley, University of California Press. 
NoRBECK, Edward and Margaret 

1956 Child training in a Japanese fishing community. In Personal character 
and cultural milieu, D. G. Haring, ed. Syracuse, Syracuse University 

Okano, M. 

1956 Shiidan kozo to personality (Group structure and personality). Shin- 
rigaku Kenkyu (Japanese Journal of Psychology) 27:8-14. 

Seligman, C. G. 

1930 Japanese temperament and character. Trans, and Proc. of the Japan 
Society. London. 28:123-142. 

Seward, G. H., ed. 

1958 Clinical studies and cultural conflict. New York, Ronald Press. [Chap- 
ters pertaining to Japan] : C. G. Babcock and W. Caudill, Personal and 
cultural factors in the treatment of a Nisei man; T. E. Bessent, An 
aging Nisei anticipates rejection; N. L. Farberow and E. S. Schneidman, 
A Nisei woman attacks by suicide; L. B. Olinger and V. S. Summers, 
The dividing path: psychocultural neurosis in a Nisei man; and M. K. 
Opler, Cultural dilemma of a Kibei youth. 

SiKKEMA, Mildred 

1947 Observations on Japanese early training. Psychiatry 10:423-432. 


1945 Psychological aspects of current Japanese and German paradoxa. Psy- 
choanalytic Review 32:73-85. 

Smith, T. C. 

1959 The agrarian origins of modern Japan. Stanford, Stanford University 




1954 Patterns of the Japanese personality indicated by the Rorschach test. 
Japanese Journal of Projective Techniques, i. 

1958 Ejiko ni tsuite — sono btnnpti to jinniigakuieki igi (Ejiko: its distri- 
bution and anthropological significance). Shonika Shinryo (Journal 
for Pediatric Practice), 21. 
SoFUE, Takao, Hiroko Sue and Taiji Murakami 

1958 Ejiko ni kans7iru bunkajinruigaktiteki kenkyil — bttmpu oyobi chiikiteki 
bent ni tsuite (Anthropological study of the E]iko, a Japanese cradle 
for child: its distribution and areal varieties). Jinruigaku Zasshi (Jour- 
nal of the Anthropological Society of Nippon) 66:jj-^i (English ab- 
stract) . 

SoFUE, Takao and Hiroshi Wagatsuma 

1959 Kokumin no shinri — Nihonjin to Obeijin (National character — Japa- 
nese, Americans, and Europeans). Tokyo, Kodansha. 

Spitzer, H. M. 

1947 Psychoanalytic approaches to the Japanese character. In Psychoanalysis 
and the social sciences. Vol. i, G. Roheim, ed. New York, International 
Universities Press. 
Stoetzel, Jean 

1955 "Without the chrysanthemum and the sword; a study of the attitudes 
of youth in post-war Japan. New York, Columbia University Press. 

Strong, E. K. 

1934 The second-generation Japanese problem. Stanford, Stanford University 

Sue, Hiroko 

1958 Ejiko ni kansnru bunkajinruigakuteki kenkyil — Miyagi-ken no ejiko 

shiyd chiiki ni okeru chosa (Anthropology study of ejiko (cradle) — 

intensive study of an ejiko-Msing community in Miyagi Prefecture). 

Jinruigaku Zasshi (Journal of the Anthropological Society of Nippon) 

66:128-136 (English abstract). 
Tsukishima, Kenzo 

1954 Noinin no paasonaritei — Kitakami-gawa churyuiki no noson no baai 
(The personality of farmers, as seen in the farming villages of the mid- 
dle reaches of the KItakami River). Tokyo University, Toyo Bunka 
Kenkyusho Kiyo (The Memoirs of the Institute for Oriental Culture) 

1955 Gyomin no paasonaritei — Nanao-wangan no gyoson no baai (The per- 
sonahty of fishermen, as seen in the fishing villages of Nanao Bay). 
Tokyo University, Toyo Bunka Kenkyusho Kiy5 (The Memoirs of the 
Institute for Oriental Culture) 7:147-190. 

1957 Nihon kdzan bnraku no ningenkankei ni kansnru bunka shinri gakuteki 
chosa hokoku (A study of human relations in Japanese mining com- 
munities from the standpoint of cultural psychology). Tokyo Univer- 
sity, Toyo Bunka Kenkyusho Kiy5 (The Memoirs of the Institute for 
Oriental Culture )i3:i49-i88. 

Wagatsuma, Hiroshi, and George De Vos 

In press "Attitudes Toward Arranged Marriage in Rural Japan," Hinnan Or- 

chapter 3 



University of Chicago 


SuBSAHARAN Africa is one of the world's great strongholds of non- 
literate peoples. Its ethnographic literature is vast, yet studies of 
culture and personality are exceedingly few. There has probably 
been less research on socialization processes, the psychodynamics of 
cultural behavior, the application of projective techniques, per- 
sonality and culture change, and culture and mental disorder in 
Africa than in any major continental area of the world. Anthro- 
pologists working there have generally eschewed such research, leav- 
ing it to psychiatrists, educators, and missionaries. The latter, many 
of whom were untrained in anthropology or scholarly research of 
any kind, have produced works which are at best straightforward 
descriptions of childhood or psychotic behavior, at worst racial 
stereotypes with scientific window dressing. All too often such 
studies concern The African Mind or The African Mentality, ig- 
noring cultural differences among Africans. We are told that TJoe 
African is impulse-driven, fear-ridden, incapable of long-range 
planning, and unable to distinguish between himself and his kin 

The case of J. C. Carothers, whose works are among the most 
widely read on the subject of mental disease and personality in 
Africa, is illustrative. Dr. Carothers is a psychiatrist who practised 
in Kenya for many years and was in charge of the Mathari Mental 
Hospital (for Africans) there. His articles have appeared in Psychi- 
atry (1948) and the Journal of Mental Science (1951) ; a mono- 
graph by him was published by the World Health Organization 
(1953), and a topical report. The Psychology of Man Man, was 

* Prepared with the assistance of National Institute of Mental Health Grant No. M-4037 (A). 



published by the Kenya Government (1954). The following quo- 
tations are characteristic of his writings. 

The native African in his culture is remarkably like the lobotomized Western 
European and in some ways like the traditional psychopath in his inability to see 
individual acts as part of a whole situation, in his frenzied anxiety and in the 
relative lack of mental ills (1951:47). 

In summary, by the nature of African experience in infancy and childhood, no 
firm foundation is laid for clear distinction of the subject and object, or for a 
proper balance in regard to those of love and hate. Tendencies to later readjust- 
ment (especially in the field of impersonal intelligence) of this distorted state 
are consistently frustrated, so that in later life there is Httle approach to a total 
personal integration, and, in dealing with any situation for which no pattern of 
behavior is prescribed by local custom, such behavior is impulsive and is marked 
by concentration on immediately presenting aspects of that situation, without 
regard for the sum of stored experience, of present perception, or of implications 
for the future (1953:107). 

If one scans the faces of the passers-by in any town in Western Europe it is 
clear that most of the people observed are impelled by some continuing inner pur- 
pose and yet are also alert to the events around them. If one leaves the ship for a 
moment at any African port, it is equally clear that most of the faces observed 
express either exclusive interest in some immediate affair or complete apathy 

In The Psychology of Man Man, in which Carothers was forced 
by the nature of the subject to consider the Kikuyu apart from 
other Africans, two factors adduced specifically to explain Kikuyu 
behavior are their "forest psychology" which comes from living 
near the edge of the forest and explains their willingness to return 
to it in Mau Mau bands, and the fact that "in Kikuyuland authority 
lacked strength" (1954:5). 

Equally ethnocentric and unscientific as the writings of Ca- 
rothers, and more Freudian, are the works of J. F. Ritchie (1943) 
and S. Davidson (1949). These authors seek to discover why The 
African is irrational, lacking in curiosity, and so forth; Ritchie, a 
school principal in Barotseland, attributes it to excessively late and 
traumatic weaning; Davidson, a psychiatrist among the Bemba, 
sees adolescent sexual promiscuity as the cause. Such analyses are 
primarily relevant, not to culture-and-personality investigations, 
but to the sociology of knowledge as examples of the use of psycho- 
logical concepts to support race prejudice. 

Although many British social anthropologists specializing in 
Africa observe what Richards has called a "psychology taboo" 
(1958:118), their field reports contain much data of interest to the 
student of culture-and-personality, particularly on family relation- 


ships, sexual behavior, the Kfe cycle of the individual, and religion. 
That they have so rarely availed themselves of psychological theory 
in the analysis of their data is perhaps attributable to the persistence 
of a tradition concerning the separation of social and psychological 
facts. Like Durkheim and Radcliffe-Brown, the present upholders 
of that tradition reject explicitly psychological explanations of 
sociocultural phenomena but often interpret field data in terms of 
individual sentiments and attitudes. Gluckman's Custom and Con- 
flict in Africa (1955) is an example of this ; much of what he terms 
conflict is equivalent to culturally patterned ambivalence within 
individuals. Another example is Nuer Religion; in it Evans- 
Pritchard makes the following comments on psychological theories 
of primitive religion: 

The psychological explanations were very varied, changing with changes in 
psychological theory. Intellectualist interpretations were succeeded by emotion- 
alist interpretations and they by psycho-analytical interpretations. Religion was 
discussed and explained in terms of association of ideas, of personification of nat- 
ural phenomena, of awe, of thrill, of fear, anxiety and frustration, of projection, 
and so forth. Most of these theories have long ago been discredited as naive 
introspective guesses (1956:312). 

In spite of this strong statement, Evans-Pritchard also rejects 
strictly sociological explanations of primitive religion and subse- 
quently concludes, "Though prayer and sacrifice are exterior ac- 
tions, Nuer religion is ultimately an interior state" (1956:322). He 
discusses the Nuer "sense of guilt" which he claims "is not just fear 
but a complex psychological state" and which "varies in intensity 
from one situation to another" (1956:31 2-3 1 3 ) . At another point 
it is stated "Nuer religious conceptions are properly speaking not 
concepts but imaginative constructions" (1956:321). This seems 
to approach a psychological view, as does his general characteriza- 
tion of Nuer religion: 

We can say that these characteristics ... of Nuer religion indicate a distinc- 
tive kind of piety which is dominated by a strong sense of dependence on God and 
confidence in him rather than in any human powers or endeavors . . . this sense 
of dependence is remarkably individualistic. It is an intimate personal relationship 
between man and God. This is apparent in Nuer ideas of sin, in their expressions 
of guilt, in their confessions, and in the dominant piacular theme of their sacri- 
fices. It is evident also in their habit of making short supplications at any time. 
This is a very noticeable trait of Nuer piety, and my conclusions are here borne 
out by Dr. Lienhardt's observations. He tells me that when he was in western 
Dinkaland he had in his household a Nuer youth whose habit of praying to God 
for aid on every occasion of difficulty greatly astonished the Dinka (1956:317- 


This description of a modal habit pattern as characteristic of a 
rehgious system is similar to one that might be written by a be- 
havioristically oriented student of personality and culture. The 
difference is that Evans-Pritchard believes that the underlying proc- 
esses are better analyzed by a theologian than a psychologist 
(1956:322). In any event it is apparent that, though personality 
theory as such is either rejected or ignored by most British Af rican- 
ists, even some of the most antipsychological of them do not over- 
look the individual and his response patterns in their ethnographic 

Culture and personality studies are not entirely missing from the 
anthropological literature on African peoples. Indeed, such studies 
can be found among the writings of some of the most eminent Af ri- 
canists — Melville J. Herskovits, S. F. Nadel, Audrey I. Richards, 
Meyer Fortes — though rarely in their best-known works. Further- 
more, a few younger scholars such as S. G. Lee combine psychological 
training with cultural sophistication to produce culture- 
personality studies of high quality. At the Fifteenth International 
Congress of Psychology at Brussels in 1957, several papers were 
presented reporting African personality research of variable qual- 
ity. The number of papers was encouraging, as was the attitude 
expressed by Dr. S. Biesheuvel of the National Institute of Person- 
nel Research, Johannesburg, in his introduction to their published 
form : 

Psychology owes a considerable debt to social anthropology for its elucidation 
of the social systems and functions that govern African community life. . . . 
Psychological research . . . should be social in its orientation, closely related to the 
work of social anthropologists, and preferably conducted on a team basis (1958a: 

In the remainder of this paper I shall outline the cultural back- 
ground to personality in Africa and then review African studies of 
culture and personality under the headings of infant experience and 
the family environment, personality development in childhood and 
adolescence, the T.A.T. in South Africa and the Congo, person- 
ality and acculturation, psychocultural interpretation of ritual, 
witchcraft, and dreams, and culture and mental disease.^ The re- 
view is not exhaustive; works of primary interest and relevance are 

^ I am indebted to my wife, Barbara B. LeVine, for an extensive search of the psychological 
literature for relevant sources, to Igor Kopytoff for bringing to my attention studies conducted in 
the Congo, and to Hans Panofsky of the African Studies Library, Northwestern University, for 
invaluable bibliographical assistance. 


discussed. Works on African intelligence (e.g., Biesheuvel 1943) 
and many psychiatric studies have necessarily been omitted. My 
intention has been to select for discussion those studies which, by 
their insights or their errors, help to point the way for future re- 

The Cultural Background to Personality in Africa 

Cultural variation among the millions of people and hundreds of 
linguistic groups in subsaharan Africa is so great as to defy any 
attempt to describe "African Culture." Ignorance of this variation 
has vitiated the attempts of many nonanthropologists to contribute 
to culture and personality studies. The culture area classifications 
ofHerskovits (1948) andMurdock (1959) provide means of com- 
prehending cultural similarities and variations at an intermediate 
level of generality between the particular culture and the entire 
continent. There are numerous cultural characteristics, however, 
which may be said to be distinctively African, although they are 
neither limited to Africa nor universal throughout it. For purposes 
of comparison with other areas of the world, I present a list of those 
distinctively African characteristics which have demonstrable or 
potential relevance to personality variables. 

1. Pastoralism. Cattle, camels, sheep, and goats are raised in 
many parts of Africa, sometimes along with agricultural activities, 
less frequently as the sole subsistence activity. A distinctive ethos or 
attitude has often been attributed to strictly pastoral and nomadic 
peoples, such as the Masai and pastoral Fulani, and to peoples such 
as the Nuer among whom pastoralism is highly valued but not ex- 
clusively practised. Herding is an important childhood occupation 
in many areas of Africa. 

2. Large and Dense Populations. African ethnolinguistic units 
tend to be large by comparison with nonliterate societies in other 
parts of the world. There are numerous African ethnic groups of 
more than a million persons (e.g., Zulu, Xhosa, Kikuyu, Ibo, Mossi) 
and many more over a quarter of a million in population; linguistic 
groups of less than 100,000 are often considered small by local 
standards. Within those large groups there is local variation in cul- 
tural practices which makes comparative studies of communities or 
districts both feasible and valuable. Population densities are high 
among many of the sedentary peoples, ranging up to 2,000 per 
square mile. In West Africa there are indigenously urban and infra- 
urban communities. 


3. Highly Developed Prestige Economy and Acquisitive Culture 
Patterns. Indigenous economic institutions are varied, but acquisi- 
tive values and status distinctions based on wealth are common 
throughout Africa. In west and west central Africa these patterns 
are related to trading and markets; on the eastern side of the con- 
tinent they most frequently involve livestock. Plural wives are al- 
most everywhere items of conspicuous consumption. 

4. Centralized Political Institictions and Institutionalized Lead- 
ership. Stateless societies outnumber centralized states in Africa, 
but the latter are found in greater abundance there than in any 
other nonliterate area of the world. Chiefs, headmen, and royal and 
aristocratic lineages play an important part in the functioning of 
many African social systems. 

5 . Unilineal Descent Groups. These are not only the most wide- 
spread form of kin group, but serve political functions in stateless 
societies and form the basis of local organization in many areas. 

6. Bridewealth. Marriage payments are customarily made to 
the family of the bride, although bride service and sister exchange 
are found in some societies. 

7. Polygyny and the Mother-child Household. Polygyny is ex- 
tremely common in Africa on the whole (see Dorjahn 1958b) and 
has important consequences for patterns of sexual behavior and 
child rearing. In many societies each wife occupies a separate house 
with her children. 

8. Initiation Kites and Genital Operations. Male and female 
initiation rites at or around puberty can be found in every major 
region of Africa, with groups lacking the rites interspersed among 
those that practise them. Circumcision and clitoridectomy are also 
widely distributed, sometimes associated with initiation, often not. 

9. Ancestor Ctdts. Beliefs and practices pertaining to ancestors 
are often associated with unilineal kin groups and are the most 
prevalent single form of African religion. The worship of nature 
deities and other gods and spirits is also found, however. 

10. Witchcraft and Sorcery. Beliefs and practices concerning 
magical aggression by humans against one another are extremely 
widespread, though their form and intensity vary. Exuvial magic 
is common. 

11. Importance of Proverbs in Folklore. In most African so- 
cieties much traditional wisdom, both moral and cynical, is sum- 
marized in proverbs which are used in everyday life and taught to 


Infant Experience and the Family Environment 

Much of what has been written on childhood in Africa by re- 
searchers and casual observers has emphasized the closeness of the 
mother-child relationship, the prolonged indulgence of infants, and 
the traumatic character of weaning. One of the few observational 
studies of African infants is that of Geber (1958), whose work was 
part of the research program organized by the International Chil- 
dren's Centre and carried out in four European cities as well as 
Africa. The children whose psychomotor development she tested 
consisted of 308 in Kampala, Uganda (cultural group unspecified 
but apparently all Ganda) , 16 in Johannesburg, South Africa, and 
30 in Dakar, Senegal (cultural groups unspecified) . The published 
conclusions do not distinguish between the groups in different parts 
of the continent but contrast them as a whole with European 

Using Gesell tests for infants past the neonate stage and methods 
devised by Andre Thomas for testing neonates, Geber found striking 
evidence of precocity in African infants. Nine-hour-old infants 
drawn into a sitting position were able to prevent their heads from 
falling back, which European children cannot do until six weeks 
after birth; two-day-olds looked at the examiner's face and seemed 
to focus their eyes, a feat not performed until eight weeks by Euro- 
pean infants. 

... up to the fifth month, the motor precocity was remarkable, especially in re- 
gard to posture. Between the fifth and seventh months, adaptivity, language and 
personal-social relations came to equal the motor development: the level was that 
of European children two or three months older (1958:186). 

Geber suggests that the initial motor precocity might be due to 
the attitude of the pregnant mother: "The arrival of a baby is al- 
ways looked forward to with great pleasure . . . and is not a source 
of anxiety. . . . The mother ... is active up to the moment of de- 
livery" (1958:194). Her "happy acceptance of motherhood may 
be related to the slight degree of tonic flexion in her new-born child" 
(1958:195). The continued precocity of older infants is attributed 
to the fact that the African children live "surrounded by affection," 
especially the "loving and warm behavior of the mothers." Geber 
states, "Before the child is weaned, the mother's whole interest is 
centered on him. She never leaves him, carries him on her back — 
often in skin-to-skin contact — wherever she goes, sleeps with him, 
feeds him on demand at all hours of the day or night, forbids him 


nothing, and never chides him" (1958:194). In support of this 
hypothesis, she cites (without specifics) the cases of some African 
children whose westernized parents kept them in cots most of the 
time and fed them on schedules; they "did not show similar precoc- 
ity after the first month, and later were inclined to be quiet and 
subdued" (1958:195). Furthermore, children examined before and 
after weaning are said to have shown "marked differences" in their 
behavior and test results; afterwards they were less lively and pre- 
cocious. This is attributed to the withdrawal of the mother's love 
and attention at the time of weaning; the Ganda custom of sending 
the child away to grandparents for months at the time of weaning 
is mentioned here. But "children for whom weaning had not caused 
a sudden break in the way of life retained their liveliness after the 
weaning, and developed without interruption" (1958:195). 

Despite the brevity of the article by Geber and its lack of detailed 
evidence, she does raise some intriguing hypotheses concerning the 
effect of desire for children, maternal love, and mother-infant con- 
tact on infant development. Like many nonanthropologists, how- 
ever, she assumes cultural uniformity for Africans, so that the 
patterns of infant care found among the Ganda of Kampala are 
generalized by implication to her South African and Senegalese 
samples. Although our knowledge of infant care in Africa is some- 
what scant, several relevant pieces of information are reliably re- 
ported and should not be overlooked: 

1. Not all African luomen desire motherhood. Among the Ila 
of Northern Rhodesia, where childbirth is followed by a 2^/2- to 
3 -year prohibition on female sexual activity, young married women 
induce abortions so that they can go on with their marital and ex- 
tramarital sexual lives. ^ Furthermore, in those groups which dis- 
approve of childbirth before marriage, the unwed mother often 
endures pregnancy in anxiety and disgrace. 

2. The close and constant relationship between mother and un- 
weaned child is jar from universal in Africa. In much of East and 
South Africa the infant is introduced to gruel within the first 
month by forcefeeding, and is left in the care of an older sibling 
during the day while the mother is working in the fields. This con- 
trasts sharply with the pattern of unbroken mother-infant contact 
found not only among the Ganda but in numerous Central African 
societies. Such variation provides the conditions for natural experi- 

*I am indebted for this information to Arthur Tuden, who did field work among the Ila m 


ments within Africa on the effects of mother-child relationships. 

3. There appears to be considerable variation among African so- 
cieties in the degree of maternal tuarmth and affection. My own 
observations indicate that there are groups in which mothers play 
with and praise their children, and others in which they ignore them 
even when ministering to their needs for nourishment and physical 
comfort. If the findings of Harlow (1958) concerning monkeys 
apply to humans, it may be that affection is a less important variable 
than the simple availability of the mother for physical contact with 
the infant. 

4. In at least some African groups mothers and other adults do 
punish and scold umveaned children for wandering too near the fire, 
for crying too much, for m^asturbation, for striking an adult, and 
so forth. The picture of total indulgence may hold for some so- 
cieties but is often exaggerated by persons who have not observed 
African families at length. 

5. Weaning from the breast is not a ''stage'' which occurs at the 
same age or with the same effects in all African societies. The mean 
ages of weaning for African groups probably range from less than 
a year to well over two years, which is a substantial segment of the 
world-wide range of variation. Evidence marshaled by Whiting 
(1954:524-525) suggests a curvilinear relationship between age at 
onset of weaning and amount of emotional disturbance the child 
shows. The greatest amount of emotional disturbance occurs in 
societies where weaning is begun between thirteen and eighteen 
months; weaning beginning before one year or over two years re- 
sults in much less emotional disturbance. This evidence is consistent 
with the common finding of traumatic weaning in African socie- 
ties, since so many of them wean in the second year of life, but it 
also suggests greater variation in amount of weaning disturbance 
(when the early- weaning and late-weaning societies are included) 
than has been recorded to date for Africans. 

Age of weaning in African societies is related to degree of poly- 
gyny, since women whose husbands have other wives tend to give 
birth at less frequent intervals (seeDorjahn 1958a) and are thereby 
able to nurse each child longer. The length of customary restrictions 
on the postpartum sexual activity of women is also involved in the 
determination of child spacing and hence often age of weaning as 
well. In some societies the postpartum taboo is justified on grounds 
of allowing the mother to devote a long time to the care of a par- 
ticular child without getting pregnant again. 


6. Methods of weaning vary among African groups. Although 
some, hke the Ganda mentioned by Geber, send children away to 
relatives to be weaned, others slap, frighten, or smear repellent sub- 
stances on the breast while keeping the child at home. 

In sum, there is variation in attitude toward motherhood, mother- 
infant contact, maternal warmth, punishment of infants, age of 
weaning, and method of weaning among African societies. They 
constitute a ready-made laboratory for the investigator who wishes 
to explore the effects of such variations on personality development. 

Albino and Thompson (1956) have carried out a study of Zulu 
weaning which could well serve as a model for future research on 
African infants. The Zulu wean suddenly, on a day set in advance, 
and this culture pattern provided the investigators with an oppor- 
tunity to observe the immediate effects of this alleged trauma on 
infant behavior. A group of sixteen Zulu infants from a single 
rural neighborhood were selected for intensive study before, during, 
and after weaning, and they were compared with a control group 
of ten urban Zulu children of roughly the same age who had been 
weaned considerably earlier. The sixteen children were given full 
nutritional examinations before and after weaning (no signs of 
marked malnutrition were found) , and they were also provided 
with a more than adequate daily ration of milk for three weeks, 
beginning a week before weaning, in order to eliminate nutritional 
discontinuity as a factor in behavior change. The mothers were 
interviewed and the children observed for seven weeks after wean- 
ing. They were tested one day before weaning, one day after, and 
one week after, on a modified Gesell Development Schedule, which 
was administered at similar intervals to the urban control group. 

The Zulu children were allowed almost unlimited access to the 
breast before their weaning, which took place at an average age of 
18.9 months by the smearing of the breast with the bitter juice of 
the aloe in their presence. When the aloes were applied to the breast, 
immediate reactions of the children took two forms: "apathetic" 
bewilderment with no attempt to run away, and running away 
from the mother without attempting to approach her again. Dur- 
ing the first few hours, only one child accepted the breast to suck 
more than once again, though most of them touched the mother's 
breasts. Negativistic, aggressive, and fretful behavior was common 
in the first two hours after weaning. In the following days, the 
child's relationship with his mother was disturbed, going through 
three distinct stages in ten of the cases: ( i ) a period of alternately 


attacking and ignoring mother, attacks occurring mainly at night 
in connection with attempts to nurse, and avoidance of mother 
occurring in the daytime; (2) a stage in which the child makes 
attempts to gain his mother's attention and to be constantly near 
her; (3) a period of increasing independence of the mother, with 
the child spending more time with other persons and showing no 
anger toward her or other signs of disturbance. 

Other changes following weaning included the following: closer 
relationship to members of the family other than mother, with 
increasing aggressiveness directed against a sibling; increasing ma- 
turity of behavior — helping in domestic tasks, imitating elders, 
speaking more distinctly with a larger vocabulary; apathy and 
anxiety during the first week, disappearing gradually thereafter; 
a marked increase in aggressive behavior, continuing in some cases 
to the end of the investigation; a marked increase in behavior disap- 
proved of by mother, such as spilling water and playing with fire; 
disturbed sleep; increase in appetite and food-demanding behavior. 
Although no change in developmental level was observed, on the 
Gesell test, the children changed from cooperative before weaning 
to negativistic or quietly uncooperative on the second administra- 
tion. This change did not occur in the control group. The authors 
conclude that weaning causes a temporary disturbance in the child's 
emotional and social life but that in the longer run it facilitates the 
development of sociability, self-reliance, and socially valued ag- 
gressiveness, and is therefore adaptive, rather than merely traumatic. 

Several anthropologists have analyzed the termination of in- 
fantile dependency on the mother in African societies by invoking 
hypotheses adapted from the psychoanalytic theory of the Oedipus 
complex. M. and F. Herskovits discovered that the classic Oedipal 
theme of killing the father "does not figure significantly in the 
corpus of Dahomean mythology," but that, "invariably, it is the 
father who initiates the hostility. His fear of eventual replacement 
by his offspring, usually made known to him through some form of 
supernatural revelation, causes him to have the son exposed, or 
killed outright." (1958:10-11). They also found sibling rivalry 
to be an important theme in Dahomean myths. They relate this to 
the life of the infant in Dahomey, who is "constantly with its 
mother" until she gives birth again; the replaced child has "a sense 
of rejection and neglect" out of which develops hostility toward the 
younger sibling. When the male grows up and becomes a father, his 
"jealousy of the son can be conceptualized as that aspect of the 


sibling rivalry complex which, through projection, reactivates the 
infantile competition for the mother in terms of competition for 
the affections of the wife" (1958:14). Thus the mythological theme 
of fathers killing their sons to avoid replacement by them is ex- 
plained as the expression of intergenerational competition which 
began "in infancy on the intragenerational level in the situation of 
sibling rivalry" (1958:1). 

A somewhat more orthodox Freudian view of the Oedipus com- 
plex provides the basic hypotheses for a cross-cultural study of male 
initiation in fifty-five societies (twelve of them African) by Whit- 
ing, Kluckhohn, and Anthony ( 1958) . They take as their cultural 
consequent the presence or absence of male initiation ceremonies 
at puberty involving painful hazing, tests of endurance and man- 
liness, seclusion from women, and genital operations. They find that 
such ceremonies are more likely to occur in societies where mother 
and infant sleep together for at least a year to the exclusion of the 
father, or where the mother is prohibited from sexual intercourse 
for at least a year after the birth of her child. In the latter case, it is 
suggested that the mother may obtain some "substitute sexual 
gratification" from nursing and caring for her infant. This intense 
relationship and/or the exclusive mother-child sleeping arrange- 
ment is seen as leading to a great emotional dependence of child on 
mother which is frustrated by the father's resumption of sexual 
relations with the mother. The child becomes hostile and envious 
toward his father, and though these feelings may be latent in child- 
hood, when the boy reaches adolescence, it is necessary for the so- 
ciety to have an initiation rite of the type mentioned above "to put 
a final stop to ( i ) his wish to return to his mother's arms and lap, 
(2) to prevent an open revolt against his father who has displaced 
him from his mother's bed, and ( 3 ) to ensure identification with 
the adult males of the society" (1958:362). 

Six African societies ( Azande, Chagga, Dahomey, Nuer, Thonga, 
Tiv) are classified as having the male initiation rite and its hypoth- 
esized childhood antecedents. While M. and F. Herskovits focus 
on the Dahomean child's replacement in the mother's affection by 
the next child, Whiting et al. go farther back to the point at which 
the mother resumes sexual relations with her husband; they view 
this as the crucial replacement, which has an impact on the child 
even before his mother is pregnant again. It is true that in Dahomey 
"the cultural ideal dictates her complete abstinence from sexual 
relations for two years at the least; a year's abstinence is still gen- 


erally observed" (Herskovits 1958:5), so that the hypothesized 
conditions for an exclusive mother-infant relationship are present, 
and this relationship is terminated in part when the period of ab- 
stinence ends. On the other hand, the custom in Dahomey is for each 
nonlactating wife to live with the husband in his house for four 
days at a time, leaving the children in her own house, so that they 
"do not witness the sexual act of their parents." This appears to 
weaken the point of Whiting et at., since they have mentioned the 
presence of the newly replaced child at the scene of parental inter- 
course by which means "the child may truly become aware of his 
replacement" (1958:362). It may be, however, that the mother's 
leaving the child for four days at a time, may function equally well 
to make him aware of his replacement and jealous of his father. The 
Herskovits hypothesis has the support of evidence from Dahomean 
culture that sibling rivalry is more important than jealousy of the 
father; the hypothesis of Whiting et al. is strengthened by the fact 
that it could "predict" patterns of childhood experience in Da- 
homey from a knowledge of its male initiation rites. 

A third analysis based on the notion of the Oedipus complex is 
that of LeVine (1959), who attempts to explain the high fre- 
quency of rape among the Gusii and the culture pattern of sado- 
masochistic heterosexuality in terms of structural and psychological 
factors. Within the Gusii family there are four kinds of parent-child 
relationships with varying degrees of sex avoidance (i.e., verbal 
and physical modesty) : father-daughter, which is strictest; father- 
son, which is next strictest; mother-son; and mother-daughter, 
respectively. The mother is more nurturant to all children than the 
father, who is rather aloof and described (by the mother) to the 
children as a strict disciplinarian. If, according to an Oedipal hy- 
pothesis, the relationship of child to cross-sex parent determines his 
later heterosexual adjustment, then we would expect the Gusii boy, 
whose mother was nurturant, to seek heterosexual experience, and 
the Gusii girl, whose father was modest in her presence and fear- 
inspiring, to fear heterosexual experience. This does not help ex- 
plain the apparently sadistic motivation of Gusii men but it is con- 
sistent with the fact that Gusii girls are more sexually inhibited 
than boys. 

Both Dahomey and Gusii, among whom father-son hostility or 
avoidance is pronounced, are patrilineal peoples. The matrilineal 
Ashanti as described by Field (i960) are characterized by an ex- 
tremely warm and affectionate father-son relationship beginning 


in infancy, while it is the mother's brother-sister's son relationship 
which involves hostility and tension: "They say that a son loves his 
father too much to kill him for the sake of inheritance, whereas 
he has no such sentiments regarding his uncle" (1960:27). This 
contrast between patrilineal and matrilineal peoples in regard to 
father-son relationships, and the role of the mother's brother in 
the matrilineal situation, constitute another confirmation of Ma- 
lino wski's assertion that the Oedipus complex is differently struc- 
tured in matrilineal societies. 

Sibling rivalry is a prominent feature of polygynous families in 
Africa; it is mentioned not only by Herskovits for Dahomey but 
by anthropologists describing many other groups (for example, 
Evans-Pritchard 1953, on the Nuer) . There is a close connection 
between sibling rivalry and the co-wife rivalry which is engendered 
by certain types of polygynous family structure. In many societies 
each wife has her own house, is allotted her own fields, and is, with 
her children, a subfamily unit operating under the more or less 
frequently exercised authority of her husband. In those groups 
which have what Gluckman ( 195 1 ) has called "the house-property 
complex," each mother-child unit is termed "a house" and is semi- 
autonomous for purposes of property holding and inheritance. 
Thus, the cattle that are paid in bridewealth for a woman's daughter 
are to be used for the marriage of one of her sons. If the family head 
decides to use the cattle to marry another wife, the bride establishes 
her "house" owing the amount of her bridewealth to the "house" 
of the older wife, and the debt should be paid from her own daugh- 
ter's bridewealth. These debts cause friction among co-wives and 
are often carried down one or two generations, causing dissension 
between the children and grandchildren of the co-wives. Further- 
more, in societies of this type, inheritance is patrilineal but the sons 
of each "house" inherit much of their wealth through their mother, 
to whose cattle and habitually used fields they have a legitimate 
claim. The more property assigned to their mother's "house" during 
the father's lifetime, the more the sons will inherit. This is also a 
factor in co-wife rivalry which becomes translated into the rivalry 
of half-brothers. In those societies where the family head appoints 
his successor or can disinherit a son, the wives vie with each other 
to have their own sons obtain paternal favor. The mother not only 
wants to see her sons prosper for their own sake and for the eleva- 
tion in status it will give her, but also because she will be dependent 
on their support in her old age. 


What impact does this "interhouse" rivalry have upon the child? 
He often grows up in a family in which relations between his mother 
and her co- wives are tense or even hostile; accusations of witch- 
craft and sorcery among the father's wives may be among his 
earliest memories. Children understand these hostile relationships 
while still young, and boys come to feel their personal stake in the 
struggle. Usually, good surface relations among half -siblings are 
maintained so as not to antagonize the family head, but there is 
likely to be considerable underlying aggression. A strong, affection- 
ate, and mutual loyalty develops between mother and sons. Mem- 
bers of the family outside his own "house," including his father, are 
likely to be viewed by the son with suspicion and treated with re- 
spect (for the father) or courtesy. This constellation of familial 
attitudes does not die easily in the individual who has acquired it. 
In many societies with segmentary patrilineages, it becomes a prin- 
ciple of social organization. When a lineage divides, it is often de- 
scendants of different wives of the founder (through their sons) 
who form the separate (and sometimes hostile) segments, which are 
frequently named after the founder's wives. In such groups the 
rivalry of co-wives and half-brothers is considered the normal pat- 
tern of social life, represented as it is in the group structure and 
mythology as well as in family interaction. 

Personality Development in Childhood and Adolescence 

There is a good deal of scattered information on the training of 
African children between weaning and puberty, but few note- 
worthy analyses. Fortes (1939) emphasizes how much the Tal- 
lensi child learns by observation without instruction by adults, and 
how a strong and early identification develops which results in spon- 
taneous imitation of adult sex role behavior, in play and, insofar as 
possible, in real life situations. Raum (1940) graphically describes 
the punitive discipline of Chagga parents. Simmons (i960) has 
provided a brief but careful description of childhood and adoles- 
cence among the Efik. In a previous work (LeVine i960) I have 
contrasted the Nuer and Gusii with respect to aggression training. 
The Nuer encourage children to fight for themselves, while the 
Gusii train their young to report quarrels and attacks to adult au- 
thority. The difference is seen as related to the greater tendency of 
the contemporary Nuer to settle quarrels by the feud, and of Gusii 
to resolve them in litigation. 

Another comparison is that by Biesheuvel (1959:11-14) of the 


Pedi and Lovedu in the Northern Transvaal. These two Bantu 
groups are closely related and similar in many aspects of culture, but 
they differ in the requirements of their social systems. The Pedi are 
warlike, group oriented, and accord a low place in society to women. 
The Lovedu, protected from attack by geographical features, are 
peace loving and individualistic, with women having high status in 
their society. Child training among the Pedi involves "frequent and 
severe corporal punishment," with the education of boys being "di- 
rected towards the development of aggressive virtues," while the 
Lovedu consider corporal punishment an "insult to personality." 
This concomitant variation, not elaborated by the author, is seen 
by him as illustrating the importance of social structure and values 
as causal factors in the socialization process. He endorses the view 
of Barry, Child, and Bacon (1959) that child rearing practices are 
adaptations to the socioeconomic environment as well as formative 
influences on the individual. 

The monograph by Read (i960) on Ngoni childhood is the first 
extended analysis of traditional African education in terms of values 
and personality. She sees values as determinants of child training 
practices, operating through an "ideal personality" or cultural self- 
image which the Ngoni aristocrats with whom she worked hold up 
to their children as a standard. Socialization is viewed as a conscious 
attempt to shape children's behavior in the direction of cultural 
ideals, and Read does not deal with unconscious processes of learn- 
ing and personality development. Her mode of analysis is illustrated 
by the following quotation. 

Two other qualities were emphasized in child training since they were expected 
of all Ngoni people in interpersonal relations. One was generosity in sharing 
anything a person had. It was a quality demanded of everyone, from the small 
child who was made to unclench his fist in which he was hiding three ground-nuts 
and give two of them to his fellows, to the big chief whose duty at a feast was to 
see that everyone had enough and to send food from his own portion to anyone 
who looked hungry (1960:155). 

An outstanding characteristic of childhood among the Ngoni 
aristocrats as described by Read is their emphasis on training in re- 
spect, obedience, and formal politeness, which is clearly related to 
the requirements of roles in their political system. I have also de- 
scribed the learning of authority relationships, contrasting child- 
hood experience among the "authoritarian" Gusii with that of the 
"egalitarian" Nuer (LeVine i960) . The entire problem of how the 
dominance-submission patterns — which are so striking a feature of 


many African political systems — are learned by individuals, de- 
serves more attention and comparative analysis than it has received. 

Some of Piaget's hypotheses concerning child development have 
been reviewed and tested on West African children by Jahoda 
(1958a, 1958b, 1958c). The most relevant of these studies is that 
of immanent justice in 120 school children of Accra, Ghana (cul- 
tural group unspecified). Jahoda criticizes the study by Havig- 
hurst and Neugarten (1953), which found that the belief in im- 
manent justice (i.e., that punishment by the physical world is an 
automatic consequence of wrongdoing) increases with age in 
Southwestern American Indian children, on the grounds that their 
scoring procedures were too "mechanical" and tended to inflate 
their results. In his Accra study, Jahoda found that "pure imma- 
nence" decreases significantly with age, naturalistic explanations 
(of accidental injuries following wrongdoing) increase signifi- 
cantly, and there is also a marked but not significant increase in 
explanations classified as "acts of god." He cites a study done in the 
Belgian Congo which found a steady decrease in punishment of 
unspecified origin ("immanent justice") , and an increase in "sim- 
ple accident" and "punishment by God." Since considerable differ- 
ence in age trends remain even when the Havighurst and Neugarten 
data are scored according to Jahoda's criteria, it would seem that 
the African children are being socialized to a different moral and 
cosmological order than Indians of the Southwestern United States. 
The paucity of cultural data in the African study allows no further 

M. H. Lystad (1960a) asked eighty-three Ashanti secondary 
school boys, aged thirteen to seventeen, to paint pictures of their 
choice. Although the pictures were analyzed primarily in terms of 
the predominance of western or traditional values, the author pro- 
poses antecedents to the form characteristics of the pictures as a 

The Ghanaian pictures are free rather than restrained in form and design. 
Ghanaian children are brought up casually. They live in an extended family set- 
ting where there is always some family member available for their needs and for 
play. Adult roles are assumed gradually as the children become more and more 
physically capable of assuming them. The relative freedom afforded them by the 
adults around them thus appears to be related to the relative freedom expressed 
in these paintings (i96oa:24i). 

This freedom is compared with the rigidity of paintings by 
French children described by Wolfenstein, who related it to the 


rigidity of behavior demanded of the French child. The correlation 
between amount of behavioral demands and amount of "freedom" 
in graphic expression, though vaguely defined, is susceptible to 
cross-cultural testing on a larger sample of cultural groups. 

Puberty rites and other initiation rituals in Africa offer a fertile 
field for culture and personality study, but relatively little has been 
done to relate such ceremonies to individual development. The 
world-wide analysis of male initiation at puberty by Whiting, 
Kluckhohn, and Anthony (1958) has been mentioned. The female 
initiation ceremony of the Bemba has been analyzed by Richards 
( 1956) , who points out the multifunctional character of such rites. 
They can be recognition of sexual and/or social maturity; they can 
sever mother-child bonds and serve as a vehicle for the expression of 
ordinarily repressed emotions by adults. Richards concludes her 
analysis by stating: 

Bemba evidence supports the suggestion that there is a correlation between 
matrilyny and girls' initiation ceremonies which emphasize the importance of 
fertility. In any society in which it is believed that women provide all the physi- 
cal substance from which the foetus is formed, this would be natural; it is the 
case in society. Moreover, the connection between matrilyny and girls' 
initiation ceremonies has been observed in a number of other African communi- 
ties. ... I have also suggested, very tentatively indeed, that in this particular 
matrilineal society there may be a connection between the lack of open hostility 
between the sexes and an unconscious feeling of guilt at robbing the man of his 
children, which is expressed in fears on the part of the women that the men will 
leave them, and on the part of the men that their wives will not respect them 
unless taught to do so by the Chisungu (initiation ceremony) (1956:160). 

In a valuable appendix, Richards surveys the literature on female 
initiation rites in Central Africa and finds that: 

. . . the correlation between girls' individual puberty rites and matrilineal 
organization is very marked in Central Africa, and that both the glorification of 
the role of the nubile girl and the praise of the man from another clan who gives 
her fertility, are consonant with the beliefs on which matrilineal organization rests 
in this area (1956:185). 

This areal survey indicating an association between matrilyny 
and girls' initiation rites in Central Africa suggests a method that 
could be used to gain a better understanding of the psychocultural 
aspects of initiation rites in many parts of Africa. In every major 
culture area, societies that have certain types of initiation cere- 
monies live right next to groups that do not have them. In western 
Kenya, for example, there are groups that have initiation and geni- 
tal operations for both sexes (e.g. Kipsigis, Gusii, Kuria) , peoples 


having initiation and genital operations for boys but not girls (Lo- 
goli and other Luhyia peoples) , and one large cultural group, the 
Luo, with no initiation or genital operations for either sex. In Ni- 
geria there are regions with an even greater variety of practises ; one 
group may perform clitoridectomy on infants without ceremony, 
while in the next group clitoridectomy may be an elaborate cere- 
monial prelude to marriage, and so forth. Such variation within 
regions provides the student of culture and personality with a labo- 
ratory. If our hypotheses linking puberty rites to child rearing on 
the one hand and social structure on the other are of any value, they 
should be able to predict from areal data on puberty rites, what dif- 
ferences in child rearing and social structure should be found in the 
area. If they cannot make valid predictions, then new hypotheses 
must be developed. In Africa there are even enough recorded in- 
stances of societies adopting and giving up initiation practices to 
enable the analyst to make comparative studies of the correlated 
factors involved in changes of this type. 

The T.A.T. in South Africa and the Congo 

Although few projective technique studies have been carried out 
in Africa, the past decade has seen at least four adaptations of the 
Thematic Apperception Test (T.A.T.) for use with African sub- 
jects, and some published studies, mostly of a methodological na- 
ture so far. Lee (1953) has designed a set of twenty- two T.A.T. 
cards (eight for each sex and six for both sexes) and published a 
manual for its use with African subjects. The subject matter for 
the pictures was based on fantasies collected from "Bantu inmates" 
of a South African mental hospital. Lee indicates that the pictures 
were originally made for use with Zulu subjects "but have since 
been found to serve their purpose adequately among Sutho, Zulu, 
Ovambo, Fingo, Xosa, Tswana, Griqua, and Swazi" (i953:pref- 
ace) , as well as among both educated and uneducated subjects. He 
recommends that the test be administered by an African, to elimi- 
nate the telling of stereotyped stories which the subjects consider 
will gain the approval of a European, and that it should be written 
or spoken in whatever language the subject finds easiest for the pur- 
pose. The procedure includes a follow-up interview, conducted a 
day or two after the test, in which the subject is asked to explain 
the sources of his plots, in particular whether they have been de- 
rived from his own experience or from folktales, myths, legends, 
books, and so forth. 


The comments by Lee on sources of the plot illustrate the impor- 
tance of knowing the history and folklore of the people to whom 
the T.A.T. is administered. For example, a certain type of folktale 
concerning a character named Cakyana is common among the 
stories told by the Zulu and Xosa subjects. If the analyst knows the 
traditional version of the story, he can interpret the idiosyncratic 
distortion (if any) which the subject has made; otherwise he con- 
founds cultural norm with individual response pattern. In another 
case Lee identifies a very dramatic plot as "the stereotyped story of 
Nongquase, the Xosa prophetess" (1953:14). One can easily imag- 
ine a psychologist who lacks knowledge of the culture and who 
has not interviewed concerning sources of the plot making incor- 
rect interpretations of such responses. 

In his manual Lee gives detailed suggestions for the analysis of the 
form and content of stories elicited by his T.A.T. cards. One of 
the most valuable sections is that giving the two most common re- 
sponses (in brief) of Zulu adults to each of twenty-two pictures. 
The author cautions that these common stories are not norms, that 
they vary considerably from one culture to another, and that par- 
ticularly noticeable differences appear with variations in age and 
degree of Westernization of subjects. Much of the content analysis 
is in terms of Murray need-press categories. A fifteen-page speci- 
men analysis of the stories of one subject (an educated Tsonga male 
from the Transvaal) is presented, using the subject's autobiograph- 
ical material and sentence completion test responses as confirmatory 
evidence for specific interpretations. One of the major findings of 
the specimen T.A.T. analysis is as follows. 

There is a certain conflict engendered . . . between the Western and tribal roles 
of the subject. His usual solution is to give unquestioning obedience (in def- 
erence) to those in authority in the tribe. This reaction is probably a reflection 
of Gilane's attitude to his father, whom he feels to be a better man than himself, 
both from the point of view of effectiveness and that of moraHty (1953:38). 

Biesheuvel, who has done a great deal of psychological testing on 
South African subjects, expresses some skepticism concerning the 
use of T.A.T. pictures. 

The rules of perspective drawing are not understood. . . . Conventional graphic 
details in the postural or facial representations of persons frequently suggested 
mutilation or blindness. Whether the latter association symbolizes the state of cul- 
tural confusion experienced by many Africans today, whether it is an expression 
of their preoccupation with a scourge which is common in Southern Africa, or 
whether it is merely a misinterpretation, at a purely perceptual level, of con- 


ventiunal pictorial cues, it is not possible to say in the present state of our knowl- 
edge concerning African perceptual habits (1958b: 176). 

One of the few publications reporting in detail the results of a 
T.A.T. study in Africa is the monograph by Ombredane (1954) . 
In a brief trip to the Belgian Congo, Ombredane administered his 
"Congo T.A.T." to twelve Basuku subjects, seven Bapende of 
Mbata-Kondo, ten Bapende of Gungu, and five workers of "varied 
races" in the town of Tshikapa. He regrets not having large enough 
samples to use statistical analysis. His analysis consists largely in 
searching for content characteristics, some of which serve to differ- 
entiate the cultural groups. For example, the Basuku often men- 
tioned food in their stories, while the Bapende, who occupy a more 
fertile and abundant environment, rarely mention food. There are 
many methodological difficulties with this study. Of the eleven 
pictures reproduced in the monograph, eight are drawings by a 
Belgian artist. The human figures are grotesquely elongated and in 
most cases suggest violent activity or macabre events; in fact they 
seem to be outpourings of aggressive impulses rather than the some- 
what ambiguous stimuli which most researchers recommend. Le- 
blanc (1958a), in a critique of the Ombredane study, reports the 
shock with which her Congolese subjects reacted to the same pic- 
tures. She criticizes Ombredane on a large number of points in- 
cluding sample size, administration, and failure to consider form 
in his analysis. 

Leblanc (1958a, 1958b, i960) has had T.A.T. pictures made up 
and has used them in a study of women in Katanga Province of the 
Belgian Congo. Her pictures, in contrast with those of Ombredane, 
were drawn by a Congolese artist and are schematic and two dimen- 
sional. The hypotheses and results of her published study will be 
discussed below; at this point the methodology is of primary inter- 
est. The tests were administered by the European researcher herself 
in Swahili, which is not the native tongue of any Congolese. This 
violates the rules laid down by Lee (1953) on two counts. Further- 
more, nowhere in her 1958 report does she mention the cultural 
groups to which her subjects belong; she refers to them as "Katan- 
gese," but the Katanga is a province containing numerous cultures. 

One of the most sophisticated contributions to T.A.T. methodol- 
ogy in Africa has been made by E. T. Sherwood, who has devoted 
a long article (1958) to the problem of designing a set of pictures 
for acculturation studies in South Africa. His own preference is 
for pictures which are structured in the sense of being aimed at par- 


ticular variables. The criteria he sets up for picture design, on the 
basis of much experimenting with different kinds of stimuh, may 
well serve as a guide to researchers in other areas. His substantive 
study was a comparison of the responses of Swazi adults who had 
been in Johannesburg less than four years with those of Swazi who 
had been there for a much longer time; it is not yet published at 
this writing.^ 

Personality and Acculturation 

More and more of the psychological studies being carried out in 
Africa have to do with acculturation. This is particularly true of 
personality testing and attitude research, in part because amount 
of education and place of residence (rural-urban) are readily avail- 
able indices of acculturation and provide the researcher with a 
source of variation on which to test hypotheses. Unfortunately, 
some investigators using acculturation as an independent variable 
have completely ignored differences in traditional culture in the 
samples being surveyed or tested. They make the bland assumption 
that more educated or urbanized Africans have more of "Western 
culture" and less of "African culture," which they characterize as 
fear-ridden or secure, restrictive or undisciplined, as they happen to 
imagine it. With the increasing number of ethnographic accounts 
of African urban life, such as Southall and Gutkind (1956) on 
Kampala, Uganda, and Longmore (1959) on sex and marriage in 
the Johannesburg metropolitan area, ignorance of the sociocultural 
context in studies of urban respondents is becoming less excusable; 
yet such studies continue to be produced. 

An example of such a study is that of Leblanc (1958b, i960) 
cited above. Subjects were drawn from the most "advanced" sec- 
tions of the Congolese populations of Elisabethville, a city of 130,- 
000 Congolese settled there since the 193 o's, and Kolwezi, a smaller 
city with 30,000 Congolese settled there since World War 11. A sen- 
tence completion test was administered to 137 subjects of both sexes 
from both cities, and a T.A.T. to 29 women from both cities. The 
sentence completion test concerned "the tribal traditional attitude 
toward women" as measured in areas of behavior "governed by a 
number of strict native customs and rituals: marital and extra- 
marital relations, sources of marital conflicts, such as aggressiveness, 
arguments about food, fecundity, sex separation, woman's infer- 

'^ An unpublished version of the Sherwood study (1961) became avnilnble too late for discussion 
in the present article. 


iority" (i958b:258). The author does not take into account the 
possibihty that the various cultural groups represented in her sam- 
ples might have different traditional attitudes toward extramarital 
relations and woman's inferiority, for example. She finds that men 
showed a more traditional attitude toward women than did women 
in her sample, and that Kolwezi subjects (less acculturated) also 
had a significantly more traditional attitude toward women than 
Elisabethville subjects (more acculturated) . 

In the T.A.T. section of the study, Leblanc selected fourteen 
women from Elisabethville and fifteen from Kolwezi, and formu- 
lated the unusual general hypothesis that acculturation would have 
the effect of ''bringing about better personality adjustment." The 
only significant differences she found between the Elisabethville 
and Kolwezi samples were greater productivity (length of stories) , 
optimism, and characterization (mentioning sex, age, and role 
characteristics of individuals as opposed to "someone" or "people") , 
in the former. If one considers that the tests were administered by a 
European in Swahili, a foreign language to all subjects, it is evident 
that most of the differences could be attributed to greater fluency 
in Swahili and more experience in contact with Europeans on the 
part of the Elisabethville women. Leblanc states that the sentence 
completion test "is a valid measure to differentiate the attitude of 
groups ... in accordance with degree of acculturation," while "the 
T.A.T. produced more doubtful results" (i958b:2 63). But she 
devises a new substantive hypothesis to explain the difference: "The 
tribal traditional attitude toward women which is unacceptable to 
the white man tends to disappear before the deeper personality var- 
iables which determine it are really modified" (i958b:2 63). Con- 
sidering the inadequacy of the research instruments employed, this 
generalization cannot be said to have received confirmation in the 

Two studies of the changing values of African students are rele- 
vant here, although they do not directly involve personality. Pow- 
dermaker (1956) analyzed the imagery in essays written by stu- 
dents of the Northern Rhodesian copperbelt; M. H. Lystad 
(1960b) analyzed, in sociological terms derived from Parsons and 
Levy, the favorite stories recounted by students in a secondary 
school outside of Accra, Ghana. In both cases, many of the students 
had been born in rural areas, and the predominance of traditional 
themes and values over urbanized western ones was a major finding 
in both studies, as it was in Lystad's (1960a) analysis of paintings 


by Ashanti schoolboys. Both the copperbelt and Accra samples, 
however, contained individuals from numerous tribal groups with 
contrasting cultures, and no attention is paid to differences in cul- 
tural background. 

In a study by Doob (1957) account is taken of cultural differ- 
ences among Africans. One of the numerous hypotheses tested con- 
cerned the relation between amount of Western education and 
deviation from traditional beliefs and practices concerning the 
family. Differences between responses of more and less educated 
groups were great (.01 level of significance) for the Zulu, weak for 
the Ganda (.10 level), and nonexistent for the Luo. This finding 
is consistent with the duration and intensity of Western influence 
in the areas in which these three cultural groups are located: Natal, 
South Africa (Zulu), Buganda, Uganda (Ganda), and Central 
Nyanza, Kenya (Luo). Doob concludes: 

Psychologically . . . the person (African) who is like a European in many re- 
spects because during or after adolescence he has learned European ways may 
resemble only superficially the person who was raised like a European in the same 
respects by his acculturated parents (1957:156). 

Thus, as in many acculturation studies outside of Africa, child- 
hood experience is seen as a crucial factor leaving persistent marks 
on the individual's response patterns. Although Doob compares 
samples of persons from differing cultural groups, he does not at- 
tempt to relate the content of traditional cultures to the attitudes 
or personality characteristics of his subjects. This remains to be done 
by students of culture and personality in Africa.^ Some of the stud- 
ies discussed in the following sections take acculturation into ac- 
count, but concentrate on interpreting culture content. 

Biesheuvel (1959) is alone in having attempted to generalize in 
broad outline about the psychological consequences of culture 
change, particularly urbanization and industrialization, in Africa. 
His analysis, which applies primarily to South Africa, is that urban- 
ization has weakened traditional African norms and sanctions with- 
out replacing them with other means of social and psychological 
control. The majority of township dwellers are portrayed as lack- 
ing the kinship bonds, the effective child training practices, and the 
conformity of traditional life, so that they are "directed only by 
impulse" and approach being "devoid of culture." This explains 
the lawlessness, violence, and laxity of sexual morals among urban 

' E. T. Sherwood (1961) has done this in his recent study of Sw.izi personnhty. 


Africans. Biesheuvel considers this similar to developments in Eu- 
rope during the dissolution of medieval society, and finds hope in 
viewing the "id-directed self" a phenomenon of transition. The 
smaller group of middle-class Africans are portrayed as closer to 
the Western values, which they learn not from their parents but 
from teachers, supervisors, and employers of European descent at 
a fairly late stage in life. Their anxiety level is high. Citing evidence 
from Rae Sherwood (1958a, 1958b) and using Riesman's typology 
of character structure, Biesheuvel concludes concerning middle- 
class Africans: 

The circumstances under which they grow up and function in the Union of 
South Africa encourage the development of the other-directed personahty type, 
in which conformity is normally regulated by anxiety. Hence it would appear 
that African personality development is proceeding straight from tradition- to 
other- direction, and that the historical stage where behavior was controlled by 
an internalized code, by guilt rather than by shame as it used to be, or by anxiety, 
as it is now, has passed them by ( 1959:18-19) . 

Turning his attention to industrialization, Biesheuvel asserts that 
traditional subsistence economies favored personal qualities which 
are not particularly adaptive for work performance in a wide range 
of industrial settings. The least westernized South African workers 
have been found to prefer the lot of a migrant mine worker because 
it allows traditionally valued leisure (even if only sporadically) 
and because of its paternalistic protection from the hazards of ur- 
ban life. However, industrial workers with a long period of urban 

... no longer look upon work as an interruption of the more meaningful and 
satisfying life of the African areas. They are committed to their daily task and 
hope to be able to advance in it. It is evident that within this group a new motiva- 
tion has made its appearance, in which the need to work is recognized as an en- 
during feature of life, capable of creating and satisfying other needs beyond the 
mere subsistence level (1959:27). 

Biesheuvel asks whether there is a unique element in the person- 
alities of Africans, and tentatively concludes he has found it in 
negritude as expounded by Leopold Senghor. 

Negrltude ... is in keeping with the concept of vitality which I consider to be 
characteristic of the behavior of African peoples. A culture in which this concept 
concerning the meaning of life reigns, can dispense with an excess of activity, 
. . . such activity is required mainly for sustained effort in pursuit of some self- 
imposed duty or goal. It has no need of the inner-directed personality structure 
which Africans are not now likely to develop to any extent, and it repudiates the 
drive element in work motivation, which is relatively lacking in Africans, as 


destructive of the main purpose of life. Though esseniially a West African creed 
and in keeping with hmitations imposed on human effort by the tropical cHmate, 
it is by no means inappropriate to certain features of African personality develop- 
ment at all cultural levels as we have found it here in the South. Indubitably, 
the philosophy of ncgritudc is far more likely to provide the black masses, in their 
transition from traditionalism, with a meaningful new culture than is provided 
by the more alien model of the West (1959:36-37). 

It would be easy to criticize this lecture by a usually rigorous psy- 
chologist for its facile generalizations, its awkward applications of 
social theory to African situations, and its occasional ethnocentrism, 
but these faults seem less important than the service he has per- 
formed by raising a number of important problems concerning the 
psychological dimension of culture change in contemporary Africa. 
Social control in urban society, adaptation to new economic cir- 
cumstances, and the development of nontraditional motives are 
problems relevant to culture and personality which are becoming 
increasingly important in the African scene. 

Psychoculturol Interpretation of Ritual, Witchcraft, and Dreams 

For many years ethnographers have been describing African cul- 
ture patterns which allow the occasional expression of feelings usu- 
ally kept strictly in check. In one of the earliest attempts to apply 
psychoanalytic theory to African data, Herskovits stated that "so- 
cially institutionalized release constitutes an outstanding character- 
istic of the Negro cultures of West Africa and of the New World" 
(1934:77) . He described the Dahomean institution of the avogan, 
the market place dance at which people are obliquely ridiculed in 
song, and the calumnious songs which co-wives sing against one 
another. Rattray was quoted to the effect that West Africans, by 
incorporating into their folklore descriptions of behavior ordinarily 
forbidden, "had discovered by themselves the truth of the psycho- 
analysts' theory of 'repression'," and "sought an outlet for what 
might otherwise have become a dangerous complex" (Herskovits 


More recent reports indicate similar phenomena in cultures of 

South Africa, Kenya, Northern Rhodesia, and Nigeria, among 
others. The institutionalized expression of ordinarily repressed hos- 
tilities and other emotions is seen by anthropologists as a safety valve, 
functional for the maintenance of institutions which require re- 
straint of individuals. Most commonly these culturally patterned 
outlets involve the expression of political hostility or antagonism 
between the sexes. In both cases, the form is frequently one of status 


reversal: the subject or vassal reprimands his chief or lord; the sub- 
missive female dons male clothes, swaggers, insults men. Gluckman 
(1955) has discussed these phenomena at length in sociological 
terms; he considers them prime illustrations of the positively func- 
tional nature of conflict in African societies. Mayer (1950) has de- 
scribed the noisy, demanding behavior of the usually obedient Gusii 
wife toward her husband at the enyangi ceremony which completes 
their marriage rites. Richards notes that in many girls' initiation 
rites "the women, who are bound to be submissive and humble to 
men at other times, are allowed to be quite outrageous in the cere- 
mony, to swagger, to shout obscenities or to attack the men" (1956: 
60) . This is true for the Gusii as it is for many other societies, but not 
among the matrilineal Bemba. Among the Nupe of Nigeria, each 
community has one of three annual ceremonies which allow "ca- 
thartic release" for impulses which are repressed in secular life 
(Nadel 1954). All three of the ceremonies concern adolescence, 
though in varying degrees, and two of them, gunnu and gani, have 
periods of sexual license as well as the imitation and caricature of 
women by boys. The third, vavu, involves (or did involve before 
it was banned by the government) an all-night battle with torches, 
sticks, and stones, between the adolescent boys of opposing village 
factions, in addition to some licentious heterosexual activity and the 
good-humored "kidnaping" and ransoming of women and old 
people by gangs of young men. Nadel uses psychoanalytic termi- 
nology in his analysis of these rituals, concluding that Nupe religion 
"in providing these outlets . . . anticipates as well as canalizes the 
working of psychological mechanisms, which might otherwise oper- 
ate in random fashion or beyond the control of society, in the 'pri- 
vate worlds' of neuroses and psychopathic fantasies" (1954:274) . 
In analyzing witchcraft and ritual, Nadel formulated his own 
version (apparently influenced by Kluckhohn's analysis of Navaho 
witchcraft) of psychoanalytic theory in relation to culture. Briefly, 
this theory is that magico-religious beliefs and practices reflect the 
anxieties and unconscious desires of a people, but that the anxieties 
and desires thus expressed have their origins in adult roles (sex and 
age roles in the context of family and kinship relations) rather than 
childhood experiences. Although it is the contemporaneous frus- 
trations and tensions of adult life which are viewed as the starting 
points, their expression in religious phenomena are discussed in terms 
of standard psychoanalytic defense mechanisms such as projection, 
displacement, and compensation. 


Nadel's position, particularly his rejection of the importance of 
child rearing, is most clearly illustrated by his comparative analysis 
of the Nupe and Gwari, two closely related tribes of Northern 
Nigeria (1952). Both have witchcraft beliefs, but the Nupe in- 
variably accuse women of witchcraft, while the Gwari accuse indi- 
viduals of both sexes. Nadel attributes this difference to the fact that 
marriage "is without serious complications and relatively tension- 
free in Gwari, but full of stress and mutual hostility in Nupe" 
(1952:21). The stress in Nupe stems from the ideal of masculine 
domination contrasted with the reality that many women are suc- 
cessful itinerant traders, usurping economic dominance in the 
family and engaging in independent behavior which is considered 
immoral. This explanation is adopted by Nadel only after he has 
searched for differences in child rearing. The only difference un- 
covered is that, among the Nupe when the two to three-year post- 
partum taboo on maternal sexual behavior is terminated, the woman 
visits her husband in his hut, leaving her children behind in her own 
hut, while the Gwari husband visits his wife so that cohabitation 
takes place in the presence of the young children. On the assumption 
that Freudian psychology would predict witnessing the primal 
scene to be the cause of sex antagonism, Nadel rejects this hypothesis 
on the grounds that the Nupe have sex antagonism but no primal 
scene, while the Gwari have the primal scene but no sex antagonism. 

An alternative explanation in terms of childhood experience is 
overlooked. It could be asserted that the Nupe child feels abandoned 
by his mother, who leaves him at night for the paternal hut, while 
the Gwari child sees his father as an intruder upon his relationship 
with the mother. One would then predict that the Nupe male 
would hate women and the Gwari male would hate men older or 
more powerful than himself. This is consistent with Nadel's state- 
ment that "Gwari informants in fact claimed that a marked hos- 
tility between father and son was a common feature of their family 
life" (1952:21). Furthermore, there is considerable evidence of 
maternal rejection among the Nupe: women practise abortion and 
use alleged contraceptives to continue their trading activities, and 
they tend to leave their children for itinerant trading when the lat- 
ter are four or five years old. Nupe women may antagonize their 
husbands by their economic activities and sexual independence, but 
they also reject motherhood and abandon their children. Thus it 
can be argued with equal cogency that the mother-child or husband- 
wife relationship is the significant antecedent to sex antagonism in 


witch beliefs. For a more crucial test than was provided by the Nupe 
and Gwari, one would need a society, or sample of societies, in which 
maternal rejection and female usurpation of male dominance in the 
conjugal relationship, were not associated. In any event, the evi- 
dence presented by Nadel is not convincing support of his rejection 
of child-rearing determinants for supernatural beliefs.^ 

In the same article on witchcraft, Nadel contrasts the Korongo 
and Mesakin, neighboring matrilineal peoples in the Nuba Moun- 
tains of the Sudan. The Korongo have no witchcraft beliefs at all; 
the Mesakin are obsessed with fears of witchcraft and frequently 
accuse each other of it, a man's mother's brother being most com- 
monly suspected. In both groups masculine vigor in youth is em- 
phasized, and at the first sporting contest after puberty there is a 
ceremony and a presentation of a gift in livestock — an "anticipated 
inheritance" — made to the youth by his mother's brother. The dif- 
ference is that among the Korongo, the gift is given spontaneously, 
while among the Mesakin, the mother's brother always refuses to 
give it at first and it often must be taken by force, a socially accepted 
procedure. Quarrels over the gift between the Mesakin youth and 
his mother's brother are frequent. If the former should fall ill, the 
latter would be suspected of witchcraft. Nadel relates this differ- 
ence to the contrasting age-class systems of the two groups: the 
Korongo have six age classes in which the valued masculine physical 
activity is gradually given up, while the Mesakin have only three 
from birth to death, so that a man relinquishes his sporting life 
abruptly at a fairly young age. For the Mesakin, "the resentment 
and refusal . . . express the older man's envy of youth and virility, 
the loss of which is brought home to him by the very request for 
the anticipated inheritance" ( 1952:26) . This resentment is allowed 
acceptable expression only in the sphere of witchcraft, and "every 
man projects his own frustrations of this nature into the allegations 
that others are guilty of witchcraft" (1952:26) . 

As in the comparison of Nupe and Gwari, so for the Korongo and 
Mesakin, Nadel examines child-rearing practices, finding them in 
this case "identical in the two tribes." He does mention, however, 
that among the Korongo premarital and highly promiscuous sex 
relations are fully accepted and openly engaged in, "while the Mesa- 
kin conceal such activity and recognize an ideal of premarital chas- 
tity." This suggests the possibility, on which Nadel makes no 

^ An extended analysis of the relation of sex antagonism to witchcraft beliefs among the Nupe, 
complete with four case histories, can be found in Nadel (1954:172—206). 


comment, that the sex training of children may differ in the two 
groups. His role analysis is again plausible, but his attempts to elimi- 
nate childhood experience as a factor are not. 

Comparing two other Nuba groups, Nadel (1955) finds that the 
religion of the Heiban is more pessimistic and fear-ridden than that 
of the Otoro. He relates this to the greater degree of order in the 
Otoro role system: wives are incorporated into their husband's lin- 
eages, adolescence is regulated in a series of stages, male homosexuals 
are allowed an accepted role as transvestites. All of these traits are 
lacking in Heiban culture, where role ambiguity is pronounced. 
Such ambiguity is seen as fostering tension which finds an outlet in 
religion. In sum, Nadel's comparative analyses are some of the most 
stimulating studies of culture and personality based on African 
material. They illustrate the advantages of taking a point of view 
wider than the single society, and the difficulties of achieving con- 
clusive results when comparing only two societies. Future students 
of culture and personality would do well to carry on the investiga- 
tions of religion and age and sex roles which he pioneered. 

In a different methodological vein, but equally concerned with 
sex and age roles, is Lee's study of Zulu dreams ( 1958) . Dreams are 
important in Zulu culture, being interpreted by diviners who fore- 
cast the future and diagnose misfortunes from them. Lee collected 
dreams from 600 Zulu men and women and made an intensive study 
of another 1 20 women to whom he administered the T. A.T., as well 
as interviewing on their dream life. He found that women reported 
a much greater amount of dream activity than men, and that the 
former dream more of intrinsically terrifying objects such as "mon- 
sters," while the latter enjoy dreaming more. In terms of central 
imagery, the number of different Zulu dreams was found to be very 
restricted. A general conclusion was that "dream content, for the 
particular sex, is derived almost exclusively from areas of social ex- 
perience permitted by the culture in the indigenous system of sanc- 
tions of some 50 to 75 years ago" (1958:270, italics in original). 
Thus, women, "acting under a very strong cultural imperative," 
dreamed of babies and children but not cattle, while men dreamed 
of cattle, their chief economic goal and source of prestige. This is 
significant since Zulu women were formerly prohibited from han- 
dling cattle, but now (in the absence of migratory-laboring hus- 
bands) do so more than men. More males also dream directly of 
fighting, which Lee interprets as related to the traditional warrior 
role of men. In his intensive study of females, Lee found that tradi- 


tional imagery and folklore were more accurate in dreams than in 
T.A.T. responses. He tentatively concludes that "the unconscious 
minds of individuals are very stable repositories of the past, and can 
be used as a valuable source of ethnographic material" (1958:280) . 
The cultural lag of the unconscious is attributed to its being ac- 
quired in childhood, vv^hile living in comparatively traditional 
circumstances and before exposure to European culture. This is 
similar to Bruner's finding in his study (1956) of acculturation in 
an American Indian group. 

In his study of Zulu females, Lee was able to obtain evidence rele- 
vant to the Freudian theory of dream interpretation. He found that 
the contents of women's dreams tend to vary with their age: young 
women dream of sex and childbearing more than older women, 
unmarried women dream of weddings more than married women. 
Many women reported dreaming of "a baby," while others men- 
tioned a recurrent dream of still water, considered by Freud and by 
Zulu diviners to symbolize childbirth. Adopting from Freud the 
hypothesis that high motivation yields directly wish -fulfilling 
dreams, while ambivalent or weaker motivation yields symbolic 
dreams, Lee compared the motivational state of women who re- 
ported baby dreams with those who mentioned still water dreams 
as more frequent. He found that baby dreams, interpreted as di- 
rectly wish-fulfilling, were more common among young married 
women "on whom the social pressure to prove their fertility is very 
great" (1958:274). Both unmarried girls, who look forward to 
childbirth but fear the social disapproval of premarital pregnancy, 
and married women with two or three children, who want to have 
more but have proved their fertility, dream of still water more fre- 
quently. Thus those who were assumed on grounds of social role to 
be more highly motivated toward childbirth had less symbolic 
dreams than those with weaker or ambivalent motivation in the 
same direction. Lee takes this as confirmation of the Freudian hy- 

Like Nadel, Marwick (1952) has related witch beliefs and ac- 
cusations to aspects of social structure which generate or direct the 
hostilities of individuals. He presents quantitative data to show that 
the Cewa, a matrilineal group of Northern Rhodesia, tend to accuse 
their own matrilineal kin of witchcraft, in contrast to the outgroup 
scapegoating found by Kluckhohn in Navaho witchcraft. Marwick 
suggests that the difference may be due to the fact that Cewa local 
groups are not as small, isolated, or crucial for subsistence as those 


of the Navaho; in fact "it may even be that among the Cewa witch- 
craft accusations have the adaptive function of being catalytic to 
the natural process of lineage segmentation" (1952:123) : 

Cewa seem to have an almost neo-Freudian recognition of the inevitable danger 
of repressing hostility for the sake of loyalty to one's close relatives. They express 
this recognition neatly by saying that members of the same matrilineage tend "to 
practice witchcraft against one another" because when they quarrel they are 
inclined "to leave unspoken words of speech with one another" (1952:217). 

Marwick interprets the Cewa data in terms of his hypothesis that 
interpersonal competition is generated by nonascriptive status re- 
lationships, that it develops into tension and conflict if the object 
competed for is intensely desired and if there are no structural means 
for regulating the competition, and that the "tension will be pro- 
jected into witch beliefs ... if there are no adequate institutionalized 
outlets for it" (1952:129). He concludes that witch beliefs and 
accusations are positively functional for the Cewa social system in 
that they destroy old social relationships, clearing away the ground 
for new ones. 

The studies reviewed in this section indicate some of the poten- 
tialities of African research for work on sex roles, sex personality, 
and the expression of culturally patterned anxieties and hostilities in 
ritual and supernatural beliefs. It is to be hoped that the excellent 
beginnings made by investigators such as Nadel and Lee will be 
followed up by systematic, comparative research into the same theo- 
retical problems. 

Mental Illness 

There is a body of psychiatric literature on Africans, much of it 
authored by psychiatrists with little anthropological sophistications 
who fail to distinguish one African cultural group from another 
and who at best make comparisons between urban and rural Afri- 
cans.*' One common finding is that "depressive" conditions are rare 
among Africans and "schizophrenic" disorders frequent, relative to 
Europe and the United States. In light of recent challenges to tra- 
ditional diagnostic categories among U.S. psychiatrists, and the 
drastic changes in psychodiagnosis which appear to be taking place, 
these older studies of Africans are of dubious value. In any event, 
there was rarely any attempt to relate the incidence or form of 
mental disease to culture patterns in specific African groups. 

* An exception is the description by Brelsford (1950) of concepts and treatment of psychopathol- 
ogy among the Bemba. 


The most relevant of the strictly psychiatric studies is that by 
Tooth ( 1950) , who surveyed mental illness in the Gold Coast (now 
Ghana) . He found a correlation between the amount of European 
contact (and traditional motives) in a region and the delusional 
content of schizophrenics in that region. 

In the North and among the "bush" peoples the delusional content was almost 
invariably concerned with the ramifications of the fetish system. The fact of 
lunacy means that an offense has been committed either against the nature spirits, 
who then trouble the offender in the form of dwarfs or fairies, or against the 
ancestral hierarchy who appear and influence the sufferer in person. Although it 
is not unusual for the insane from this section of the population to speak of them- 
selves as under the control of God, no example was found of identification with 
the Deity. It is possible because of the concentration of missionary activity in 
the South that the identification of the insane with an anthropomorphic God 
is so common there. Messianic delusions were not met with outside the asylum, 
where identification with Christ was sometimes combined with one or more of 
the leading figures of international politics. Delusions of grandeur were not found 
among the "bush" people but among the insane in Ashanti, delusions of great 
wealth were common and often associated with claims to royal birth and con- 
nections with powerful chiefs. It was only in the more sophisticated South that 
living individuals or groups, usually connected with the government and operat- 
ing by means of electricity, wireless or television, took precedence in the delu- 
sions of the insane over the traditional supernatural agencies (1950:52). 

Tooth hypothesizes that the situations of personal choice intro- 
duced under Westernization lead to mental disorder, but he is unable 
to find quantitative evidence of more psychosis among Westernized 
segments of the population. With respect to treatment of psy- 
chotics, he mentions the frequent sight of them at market places 
(a possible locus for ethnopsychiatric field work! ) and contrasts the 
attitude toward psychosis in three regions of the country. He con- 
cludes that "the Africans have evolved a system which cares for 
quite 80 per cent of their insane under conditions which compare 
favorably with those provided by the European authorities" 

Among the few studies of culture and mental disease carried out 
in Africa are those on related Nguni groups by Laubscher (1937) 
and Lee (1950). Laubscher, a psychiatrist who did field work 
among the Tembu and related Fingo of South Africa, describes the 
role of mythical beings in their traditional explanations of psy- 
chotic behavior, in the delusions of hospitalized psychotics, and in 
the dreams of normals. The beings include hypersexual dwarfs, 
blood-eating and hypersexual birds, and snakes harbored in the fe- 
male organs. In many cases these creatures are viewed as gratifying 


the extramarital sexual cravings of females. The imagery itself and 
interpretations of it by diviners are so suggestive of Freudian con- 
cepts that one might almost say that psychoanalytic theory is part 
of the Tembu-Fingo belief system. 

Lee has analyzed almost identical phenomena among the Zulu, 
a closely related Nguni people, and his analysis is freer of a heavy- 
handed early Freudianism than that of Laubscher. He describes the 
syndrome known as "Bantu disease" or ufiifunyana, which is recog- 
nized by the Zulu as a nonorganic condition similar to the state of 
possession manifested by a "witch doctor" during his apprenticeship 
(This is true of the Tembu and Fingo as well). 

Stereotyped dreams involving the above-mentioned supernatural 
beings and, among present-day Zulu, involving Indians (many of 
whom live in Natal) , are an integral part of the syndrome. The dis- 
ease is most commonly found among women, who complain of 
pains in their lower abdomen, sometimes develop paralysis, and also 
have seizures during which they talk incoherently in what their 
neighbors assume to be an Indian language. The women often dream 
of "tokoloshe," the bearded dwarf with a huge penis, and believe 
that he rapes them at night. Lee describes three rather different 
cases of nftifunyana, two of which he considers "pure cases of con- 
version hysteria." The women afflicted suffer from obvious sexual 
fears and frustrations and their disorders were precipitated by sexual 
crisis, in one case desertion by a husband, in the other a threat by a 
rebuffed lover. The third case, that of an old man, appeared to be 
related to sexual jealousy. Lee states his conviction that this disease, 
its high frequency among women, and its apparent increase in re- 
cent years, are related to the "heavy anxiety load" of Zulu culture, 
and indicates that he will carry out further studies "directed at the 
discovering of specific reasons for the obviously insecure personality 
pattern which seems to be so common among the Zulu people" 
(1950:18). Loudon (i960) has speculated on the correlates of 
ufufunyana, but without any convincing evidence. 

Nadel (1946) has attempted to relate shamanism among the 
Nyima and other peoples of the Nuba mountains in the Sudan to the 
incidence of mental disease among them. The shamanism he de- 
scribes is similar to that found in Central Asia, is highly institu- 
tionalized, and plays an important part in the medical and religious 
aspects of Nyima culture. The shaman must be capable of spirit 
possession which is similar in overt behavior to cataleptic seizures; 
instances of possession observed by Nadel appeared to him to vary 


in their degree of "sincerity" and conscious fakery, but he was 
convinced that a majority of them resulted in seizures over which 
the shaman had Httle conscious control. "Insanity" (not defined) is 
said to be rare among the Nyima, but epilepsy is widespread (esti- 
mated at one in loo) and is recognized as frequent by the people 
themselves and by medical officers in the district. Epilepsy is not re- 
garded by the Nyima as spirit possesssion or as a qualification for the 
role of shaman, but six of the shamans interviewed had epilepsy in 
their families and, of eight hereditary shamans, only two claimed 
that none of their relatives had been epileptic. The shamans them- 
selves are not epileptics and have no histories of mentally deranged 

The possibility is considered that shamanism in the Nuba moun- 
tains is associated with a low incidence of "insanity" and a high 
incidence of epilepsy. However, the Dilling, who also have shaman- 
ism, are estimated to have a relatively high incidence of insanity 
(one in 300) but little or no epilepsy; the Koalib, another shamanis- 
tic group, have much less of both conditions (one in 500 for in- 
sanity, one in 1000 for epilepsy). Furthermore, nonshamanistic 
Nuba groups have incidences of insanity both lower and higher 
than those estimated for the shamanistic groups. Thus this simple 
hypothesis is rejected. Taking into account the fact that shamanism 
is increasing in intensity and frequency among shamanistic groups 
and is also spreading to nonshamanistic groups, Nadel suggests its 
relationship to the "psychologically unsettling" impact of culture 
change brought about by contacts with Western civilization. This 
change "among the Nyima as among all primitive communities . . . 
must create and foster emotional instability, neurotic and hysterical 
leanings, that is, the constitutional qualifications of a shaman" 
(1946:36) . The hypothesis is formulated that shamanism is a pre- 
ventive measure for mental health: 

Shamanism still leaves in existence and without a social "niche," the deviant 
and abnormal personality, though the borderline between normal and abnormal 
differs from that valid in non-shamanistic groups. But it remains an open ques- 
tion whether shamanism does not in a different sense "absorb" mental derange- 
ment; the institutionalized catharsis which it offers may well have the therapeutic 
effect of stabilizing hysteria and related psycho-neuroses, thus reducing a psycho- 
pathic incidence which should otherwise be much larger (1946:36). 

Thus the shamanistic groups may be able to cope with the general 
psychological disturbance resulting from acculturation without a 
higher incidence of mental disease. In other words, Nadel rejects a 


synchronic hypothesis, that shamanistic and nonshamanistic groups 
differ in their incidence of mental disease, in favor of a diachronic 
hypothesis to the effect that the groups will differ in the amount of 
increment in mental disease under changing conditions. 

Thus the hypothesis I suggest is verifiable ... for if it is true, it must be pos- 
sible to show that psychoses and kindred disorders are increasing among the non- 
shamanistic groups, while in the shamanistic groups the increase of shamanism 
would go hand in hand with a relatively undisturbed mental stabihty ( 1946:37) . 

Nadel did not have the data to test this hypothesis, but his study 
provides an excellent example of research design for future students 
of the relation between culture patterns and mental disease in 
changing African societies/ 

Messing (1958, i960) has analyzed the Zar spirit-possession cult 
of the Amhara of Ethiopia as group psychotherapy for a wide range 
of emotional disturbances ''ranging from frustrated status ambi- 
tion to actual mental illness." Married women are the most frequent 
patients, and the cult functions not only to mitigate symptoms, 
but also to provide a group context in which deviants are reinte- 
grated into society. The sexual symbolism of the relation between 
the patient and his Zar, and the manner in which the cult reflects 
Ethiopian social stratification, are some of the fascinating aspects 
of the study. Spirit-possession phenomena of a similar type occur in 
West Africa and the Caribbean; their comparative analysis in so- 
ciopsychological terms would contribute greatly to our understand- 
ing of the psychiatric functions of religion. 

The most thorough investigation of mental illness in a single Afri- 
can culture is that by Field (i960) among the rural Ashanti. Utiliz- 
ing her previous experience as an ethnographer. Field returned to 
the Ashanti as a psychiatrist and set herself up near a shrine where 
troubled people come to receive help from a deity whose priest be- 
comes possessed and communicates advice from the god. It was pos- 
sible for her to observe and obtain case histories on those supplicants 
who were mentally ill, and she conducted some local surveys as well. 
The troubles and desires which normal people bring to the shrine are 
described in detail before the psychiatric data are presented. For the 
most part, standard diagnostic categories are used, and the emphasis 
is on similarities between behavior patterns observed in the field and 
those found among Europeans. 

One of Field's findings illustrates perfectly the need for intensive 

^Another excellent example is provided by Scotch (i960) in relating essential hypertension 
to changes accompanying urbanization in a quantitative study of rural and urban Zulu. 


community study outside the mental hospital to get a valid picture 
of the incidence of various mental disorders in a given population. 
As mentioned above, the older psychiatric studies (including that 
of Tooth, who worked in Ghana) are unanimous in stating that de- 
pression is extremely rare, and they present quantitative data to 
prove it. However, Field states: 

Depression is the commonest mental illness of Akan rural women and nearly 
all such patients come to the shrines with spontaneous self-accusations of witch- 
craft. . . . The depressive personality is, in sickness and health, self-effacing and 
is seldom a disturbing nuisance. She is therefore the last type of patient who would 
ever find her way to any kind of European hospital unless she had some concurrent 
and conspicuous physical trouble. ... It is not surprising therefore that psychia- 
trists and other doctors who see patients only in hospitals and clinics should have 
the idea that depression in Africa hardly exists ( 1960:149) . 

This discovery of depressive disorders is an important one and is 
adequately documented in the case histories, but there is no discus- 
sion of the psychocultural determinants of guilt in Akan individ- 
uals. In fact, Field appears to regard the guilt and depression as a 
tendency not produced by the conditions of Akan culture but oc- 
curring equally among all peoples who actively believe in witch- 
craft. She claims that only the confessions of depressives can keep 
such beliefs alive in a group; the fantasies of paranoids are not suffi- 
cient. This is contrary to fact, for there are numerous African so- 
cieties in which witchcraft is a major preoccupation but no one ever 
confesses to being a witch. Field does not take into account the vari- 
ation of witch beliefs among African societies, and this leads her 
away from investigating the peculiar conditions in Ashanti which 
make confession a pronounced pattern. 

Cultural norms are considered in the section on paranoid reac- 
tions: "In a country where nobody looks twice at a lorry announc- 
ing in big letters, 'Enemies all about me,' or 'Siiro nnipa' (Be afraid 
of people) , it is clear that our ideas of what constitutes a morbidly 
paranoid attitude must be revised" ( 1960:296) . Nevertheless, Field 
asserts that it is quite possible to distinguish the controlled paranoia 
of the normal Ashanti from abnormal paranoid reactions. The 
valuable contribution of the study is that it presents psychotic be- 
havior in cultural context, with the element of supernatural belief, 
which is so important in these disorders, clearly delineated in its re- 
lation to precipitating social circumstances and organic factors. 
The etiology of the psychoses described is considered as being outside 
the limits of the study, in part (I suspect) because the author be- 


lieves that the Akan do not diflF er significantly from other peoples in 
their mental illnesses but only in the cultural forms which these dis- 
orders take. 

Although Field's monograph contains the largest number of pub- 
lished psychiatric case histories from a single African group, it 
should be noted that Tooth (1950) also includes numerous case his- 
tories, and Sachs (1947) did a book-length case history of a Johan- 
nesburg witch doctor. The study by Bohannon (i960) of homicide 
and suicide in seven African societies, although it is not a psychologi- 
cal analysis, does contain case histories and is important as the first 
comparative study of deviant behavior in Africa. 


The foregoing survey bears out the initial assertion that relatively 
little culture-and-personality research has been carried out in Af- 
rica. In fact, considering how little has been done, it is remarkable 
that there are studies of the quality of Albino and Thompson 
(1956) on weaning, Lee (1950, 1953, 1958) on adult personality 
and projective techniques, and Field (i960) on mental illness. The 
still untapped and largely unrecognized potentialities of Africa as a 
field for culture and personality study necessitate attention to the 
possible lines of future research. In the recommendations which fol- 
low, emphasis is placed on types of research which utilize the pecul- 
iar advantages of Africa as a major ethnographic area. Thus the 
large number of distinct ethnic groups suggests the feasibility and 
importance of comparative studies; the vast accumulation of pub- 
lished ethnographic material, particularly on social organization, 
makes analysis of existing literature valuable, with personality and 
social structure a natural emphasis; the recency of Western contact 
in many groups is conducive to studies of culture change, and the 
differential exposure to Western culture of persons with the same 
traditional culture (in the rural-urban and educated-uneducated 
dichotomies) makes controlled comparisons possible; the migrant 
labor situation creates the conditions for studies of the impact of 
absent fathers on personality development; variations in the pres- 
ence and content of initiation rites present themselves as a problem 
for psychocultural investigation, and so forth. 

Comparative Analysis of Existing EtlmograpJoic Materials. No 
area of the world has as much reliable information on social organi- 
zation in as many different societies. Correlational studies of per- 
sonality and social structure could include relationships between 


family roles and sociopolitical roles, between the economic and so- 
cial position of women and mother-child relationships, among dif- 
ferent forms of culturally patterned aggression, such as warfare, the 
feud, sorcery, and so forth, and between sex roles and patterns of 
sexual behavior. 

Comparative Socialization Studies. We need basic material simi- 
larly collected on a large number of traditional African cultures. 
The kinds of data required range from motor development and 
infant nutrition through parent-child and sibling relationships to 
the socialization of sex, aggression, and dependence, and training 
in achievement, responsibility, and skills. Only by the collection of 
comparable materials on traditional child-rearing patterns will it 
be possible to find the conditions under which traditional culture 
patterns were learned and adapted to individual needs, and to estab- 
lish baselines for studies relating to socialization and culture change. 
Feasible studies of special significance include: (i) the effect of 
structural variations (different polygynous arrangements, virilocal 
versus uxorilocal marriage, more and less authoritarian extended 
family patterns, varying divorce rates, high and low status position 
of women) on child experience and behavior; (2) the effect of eco- 
nomic factors (pastoral versus agricultural subsistence patterns, 
differentiated versus undifferentiated economic role systems) on 
child experience and behavior; ( 3 ) the effect of mother-child sepa- 
ration (at termination of postpartum sexual taboo, weaning, or re- 
placement by a sibling) on children conditioned to varying amounts 
of initial nurturance by mother, with dependency weaning varying 
in its abruptness from one group to another; (4) the effect of dif- 
ferential severity of sex and aggression training on cultural behavior 
in those motivation systems; (5) the differing courses of adolescent 
development in cultures with and without male and female initia- 
tion rites at puberty; (6) the connection between varying political 
values (for example, authoritarian versus egalitarian) and the values 
concerning interpersonal behavior which are transmitted to chil- 

Urbanization and Education. These two processes are of funda- 
mental importance in contemporary culture change in Africa and 
can be expected to have their correlates in personality change. In 
one kind of research design, urban and ural, or educated and unedu- 
cated, individuals belonging to the same ethnic group can be com- 
pared on indices of culture stress (mental illness, psychosomatic 
disorders, suicide, crime) , patterns of child rearing, and values con- 


cerning interpersonal relations, supernatural phenomena, political 
behavior, achievement, and ethnic parochialism. Another approach, 
increasingly feasible under contemporary conditions, is to study 
differences on these variables between the first urban or Western- 
educated generation and later generations whose parents have been 
urbanites or educated persons. The varying reactions of different 
cultural groups to the same urban or school environment provide 
another possibility for personality study, with the emphasis on the 
extent to which traditional behavior patterns are persisting under 
changed social conditions. The effect of labor migration on child 
experience and identification processes, adolescent adjustment in 
urban settings, and changes in female roles brought about by eco- 
nomic development, are examples of specific topics which deserve 

Comparative 'Psychiatry. Primary attention must be paid to the 
collection of basic data concerning the incidence of mental illnesses 
of various types and their cultural contexts, in variety of African 
populations. This is a tremendous task in itself, and will necessarily 
involve medical investigators to distinguish functional disorders 
from the behavioral effects of trypanosomiasis and nutritional defi- 
ciencies, as well as anthropologically sophisticated personnel to con- 
centrate on cultural reactions to behavioral deviance. Some special 
problems which the African studies to date suggest include: the 
development of sexual disorders such as impotence and conversion 
hysteria in societies which set a high value on fecundity, but which 
vary in the requirements of their sex roles; the particular relation 
of cultural stresses affecting women to their development of de- 
pressive conditions (as in Ashanti) and various forms of dissociative 
behavior which involve spirit possession as a psychotherapeutic 
technique; the differential incidence of mental illnesses in West- 
ernized and non- Westernized segments of the population (men- 
tioned above) ; the relation of sorcery and witchcraft beliefs to 
paranoid conditions. 

In making these suggestions I have avoided suggesting particular 
techniques to be used. I assume that investigators will choose be- 
havioral observation, projective techniques, interviews, question- 
naires, dreams, or life histories, according to the problem under 
study and their own assessment of the validity and reliability of 
these research instruments. 

In the long run, systematic studies of culture and personality in 
Africa will benefit not only this developing subdiscipline but also 


the new nations of Africa in their attempts to modernize themselves 
while meeting the needs of their culturally heterogeneous popula- 
tions. This difficult task cannot be accomplished without an under- 
standing of the behavior patterns and motivations of the changing 
but still mainly traditional ethnic groups within their borders. 


Albino, Ronald C. and V. J. Thompson 

1956 The effects of sudden weaning on Zulu children. British Journal of 
Medical Psychology 29:177—210. 
Barry, H., I. L. Child, and M. K. Bacon 

1959 Relation of child training to subsistence economy. American Anthro- 
pologist 61:51-63. 
BiESHEuvEL, Simon 

1943 African intelligence. Johannesburg, South African Institute of Race 

1958a Objectives and methods of African psychological research. Journal of 

Social Psychology 47:161—168. 
1958b Methodology in the study of attitudes of Africans. Journal of Social 

Psychology 47:169—184. 
1959 Race, culture and personality: The Hoernle Memorial Lecture 1959. 
Johannesburg, South African Institute of Race Relations. 
Bohannon, Paul (ed.) 

i960 African homicide and suicide. Princeton, N.J., Princeton University 

Brelsford, W. V. 

1950 Insanity among the Bemba of Northern Rhodesia. Africa 20:46-54. 
Bruner, Edward M. 

1956 Cultural transmission and cultural change. Southwestern Journal of 
Anthropology 12:191—199. 
Carothers, J. C. 

1948 A study of mental derangement in Africans and an attempt to explain 
its peculiarities, more especially in relation to the African attitude to 
life. Psychiatry 11:47—86. 

195 1 Frontal lobe function and the African. Journal of Mental Science 

1953 The African mind in health and disease, a study in ethnopsychiatry. I 
Geneva, World Health Organization Monograph Series, No. 17. 

1954 The psychology of Mau Mau. Nairobi: The Government Printer, Colony 
and Protectorate of Kenya. 

Davidson, S. 

1949 Psychiatric work among the Bemba. Rhodes-Livingstone Journal 7:75— 


DooB, Leonard W. 

1957 An introduction to the study of acculturation. Journal of Social Psy- 
chology 45:144—160. 

DoRjAHN, Vernon R. 

1958a Fertility, polygyny and their interrelations in Temne society. Ameri- 
can Anthropologist 60:838-860. 
1958b The factor of polygyny in African demography. In Continuity and 
change in African cultures, W. Bascom and M, Herskovits, eds., Chi- 
cago, University of Chicago Press. 
Evans-Pritchard, E. E. 

1953 Kinship and marriage among the Nuer. London, Oxford University 

1956 Nuer Religion. London, Oxford University Press. 
Field, M. J. 

i960 Search for security: an ethno-psychiatric study of rural Ghana. 
Evanston, 111., Northwestern University Press. 
Fortes, Meyer 

1939 Social and psychological aspects of education in Taleland. London, 
International African Institute Memorandum. XVII. 
Geber, Marcelle 

1958 The psycho-motor development of African children in the first year, 
and the influence of maternal behavior. Journal of Social Psychology 

Gluckman, Max 

195 1 Kinship and marriage among the Zulu of Natal and the Lozi of North- 
ern Rhodesia. In African systems of kinship and marriage, A. R. Rad- 
cliffe-Brown and C. D. Forde, eds., London, Oxford University Press. 
1955 Custom and conflict in Africa. Glencoe, The Free Press. 
Harlow, Harry 

1958 The nature of love. The American Psychologist 13:673-685. 
Herskovits, Melville, J. 

1934 Freudian mechanisms in primitive Negro psychology. In E. E. Evans- 
Pritchard, Raymond Firth, Bronislaw Malinowski, and Isaac Schapera, 
Essays presented to C. G. Seligman. London, Kegan, Paul, Trench, 
Trubner and Co. 
1948 Man and his works. New York, Alfred A. Knopf and Co. 
Herskovits, M. J. and Frances Herskovits 

1958 Sibling rivalry, the Oedipus complex, and myth. Journal of American 
Folklore 71:1—15. 
Jahoda, Gustav 

1958a Child animism: I. A critical survey of cross-cultural research. Journal 

of Social Psychology 47:197-212. 
1958b Child animism: II. A study in West Africa. Journal of Social Psychol- 
ogy 47:213—222. 
1958c Immanent justice among "West African children. Journal of Social 
Psychology 47:241-248. 


Laubscher, B. J. F. 

1937 Sex, custom, and psychopathology, a study of South African pagan 
natives. London, Routledge. 

Leblanc, Maria 

1958a La problematique d'adaptation du T.A.T. au Congo. Zaire 12:339-348. 

1958b Acculturation of attitude and personality among Katangese women. 
Journal of Social Psychology 47:257-264. 

i960 Personnalite de la femme Katangaise Louvain Publications Universitaire. 
Lee, S. G. 

1950 Some Zulu concepts of psychogenic disorder. Journal for Social Re- 
search, Pretoria, 1:9—18. 

1953 Manual of a thematic apperception test for African subjects. Pieter- 
maritzburg, University of Natal Press. 

1958 Social influences in Zulu dreaming. Journal of Social Psychology 47: 

LeVine, Robert A. 

1959a Gusii sex offenses: a study in social control. American Anthropologist 

61 :965-990. 
i960 The internalization of political values in stateless societies. Human Or- 
ganization 19:51-58. 
LoNGMORjE, Laura 

1959 The dispossessed: a study of the sex-life of Bantu women in and around 
Johannesburg. London, Jonathan Cape. 

Loudon, J. B. 

i960 Psychogenic disorder and social conflict among the Zulu. In Culture 
and mental health, M. K. Opler, ed. New York, Macmillan. 

Lystad, Mary Hanemann 

1960a Paintings of Ghanaian children. Africa 30:238—242. 

1960b Traditional values of Ghanaian children. American Anthropologist 
Marwick, M. G. 

1952 The social context of Cewa witch beliefs. Africa 22:120—35, 215—33. 

Mayer, Philip 

1950 Privileged obstruction of marriage rites among the Gusii. Africa 20: 

Messing, Simon D. 

1958 Group therapy and social status in the Zar cult of Ethiopia. American 
Anthropologist. 60: 1 1 20—1 126. 

i960 Group therapy and social status in the Zar cult of Ethiopia. In Culture 
and Mental Health, M. K. Opler, ed. New York, Macmillan. 

MuRDocK, George Peter 

1959 Africa, its peoples and their cultural history. New York, Macmillan. 

Nadel, S. F. 

1937a Experiments on culture psychology. Africa 10:421-435. 
1937b Field experiments in racial psychology. British Journal of Psychology 
28: 195-21 1. 


1946 A study of shamanism in the Nuba mountains. Journal of the Royal 
Anthropological Institute 76:25—37. 

1952 Witchcraft in four African societies: an essay in comparison. Ameri- 
can Anthropologist 54:18-29. 

1954 Nupe religion. London, Routledge & Kegan Paul Ltd. 

1955 Two Nuba religions: an essay in comparison. American Anthropologist 

Ombredane, Andre 

1954 L'exploration de la mentalite des noirs congolais a moyen d'une epreuve 
projective: Le Congo T.A.T. Memoire d'Institut Royal Colonial Beige, 


1956 Social change through imagery and values of teen-age Africans in 
Northern Rhodesia. American Anthropologist 58:783—813. 

Raum, O. F. 

1940 Chagga childhood. London, Oxford University Press. 
Read, Margaret 

i960 Children of their fathers; growing up among the Ngoni of Nyasaland. 
New Haven, Yale University Press. 
Richards, Audrey L 

1956 Chisungu: a girl's initiation ceremony among the Bemba of Northern 
Rhodesia. London, Faber & Faber. 

1958 Review of Custom and Conflict in Africa by Max Gluckman. Man 
Ritchie, J. F. 

1943 The African as suckling and as adult (a psychological study). Living- 
stone, The Rhodes-Livingstone Papers, No. 9. 
Sachs, "Wulf 

1947 Black Hamlet. Boston, Little, Brown. 

Scotch, Norman A. 

i960 A preliminary report on the relation of sociocultural factors to hyper- 
tension among the Zulu. Annals of the New York Academy of Sciences 

Sherwood, Edward T. 

1957 On the designing of T.A.T. pictures, with special references to a set 
for an African people assimilating Western culture. Journal of Social 
Psychology 45:162—190. 

1 96 1 Swazi personality and the assimilation of western culture. Unpublished 
Ph.D. dissertation, University of Chicago. 

Sherwood, Rae 

1958a The Bantu civil servant. UnpubUshed report to the National Council 

for Social Research, South Africa. 
1958b The Bantu clerk: a study of role expectations. Journal of Social Psychol- 
ogy 47:285-316. 
Simmons, Donald C. 

i960 Sexual life, marriage, and childhood among the Efik. Africa 30:153- 


SouTHALL, A. W and P. C. W. Gutkind 

1956 Townsmen in the making. East African Studies No. 9. Kampala, 
Uganda, East African Institute of Social Research. 
Tooth, Geoffrey 

1950 Studies in mental illness in the Gold Coast. London, His Majesty's Sta- 
tionery Office, Colonial Research Publication No. 6. 
Whiting, John W. M. 

1954 The cross-cultural method. In Handbook of social psychology, Gard- 
ner Lindzey, ed. Vol. i. Cambridge, Mass., Addison-Wesley. 
Whiting, John W. M., R. Kluckhohn, and A. Anthony 

1958 The function of male initiation ceremonies at puberty. In Readings 
in social psychology (3d edition) , E. E. Maccoby, T. M. Newcomb, and 
E. L. Hartley, eds. New York, Henry Holt and Co. 

chapter 4 



University of North Carolina 


The distinction between ethnology or cultural anthropology and 
that subdiscipline of anthropology, culture and personality, rests 
on which of two ideally distinct points of view an observer adopts. 
Paraphrasing Sapir (1932; cf. Kluckhohn 1944:602-604), an 
ethnologist looks at a segment of behavior as a culture pattern, while 
the student of culture and personality studies the same segment 
from the standpoint of the persons whom it directly involves. The 
behavior has "person-defining value." Using other words, in culture 
and personality an observer focuses on the subjective side of culture, 
that is, culture as experienced or manifested by a composite (or 
typical) individual — the Hopi child, the Sioux Indian, or the U.S. 
American. Or an observer studies a real individual or categories of 
people to see how they experience a way of life. Culture and per- 
sonality implies sustained concentration on the explicit and implicit 
meanings which cultural traits (artifacts, ceremonies, legal norms, 
or epic poems) possess for persons in the community. In a somewhat 
different approach, demonstrated by Ruth Benedict (1932:24), the 
student of culture and personality may choose to see culture as the 
personality of its carriers writ large. True, all cultural anthropology 
gives attention to persons, meanings, and to the subjective. In cul- 
ture and personality there is simply more emphatic or explicit recog- 
nition of the social actor as a person, often to the relative exclusion 
of social structure, technology, ideological systems, and historical 

* Several years ago with the assistance of Lewis Binford and under the auspices of tlie Insti- 
tute for Research in Social Science, University of North Carolina, I gathered material for a 
history of culture and personality. A portion of that material has been used for this essay. 
I am grateful to the Institute for assistance in preparing the present work for publication; a 
brief version was presented at the 1958 meeting of the American Anthropological Association. 



background. In a work of culture and personality, whether it is a 
record of children's development, a life history, or an interpretation 
of Rorschach tests, the individual looms very large. 

My object in this chapter is to review culture and personality 
research which has been conducted in native North America and, 
though somewhat more incidentally, below the border. I do not 
propose to write an all-inclusive history but, rather, a judicious 
record of general accomplishments. My emphasis will be on the 
bench marks which reveal new interests, methods, or levels of so- 
phistication. The chapter is divided into three parts: an introduc- 
tion, which is herewith concluded; an evaluative review of research 
on North American Indians; and a final section of assessment and 
discussion. In this last part I realize that I go beyond reviewing 
North American Indian studies. 

Review and Evaluation 

Aboriginally North America was a continent of varied lifeways, 
traces of which still remain. Practically all over the continent, how- 
ever, missions, schools, traders, and government administrators have 
churned up culture change. The displacement of war, hunting, and 
ceremonies brought about a profound alteration in the traditional 
roles of men and women and in all other interaction patterns. The 
socially standardized milieux in which children were aboriginally 
socialized have been substantially transformed. In the United States, 
as well as in southern Canada, Indians cluster on reserves and oc- 
cupy a special status as far as the larger community is concerned. 
Someone might regard these conditions as evidence that the Ameri- 
can Indians, with a few exceptions, no longer possess truly exotic 
cultures. He might believe that the Indians could hardly be worth- 
while subjects to study in order to learn something about the diverse 
systems of personality that occur under differing cultural condi- 
tions. He might believe that, while the Indians who live under reser- 
vation conditions might at best reveal traumatized personalities, 
casualties of culture change, they will not provide the kinds of in- 
sights that it is possible to obtain, say, in parts of Africa and the 
Southwest Pacific. One could conceivably interpret some of the 
works to be reviewed in this chapter as supporting such extreme 
expectations. Anthropologists have indeed found some Indian social 
personalities to be laden with conflict and uncertainty. But it is 
worth remembering that the theoretical point from which much 
culture and personality research departs has been almost deliber- 


ately concerned with discovering pathology in people's world- and 
self-views. Anthropologists who employed a crisis-oriented ap- 
proach when they studied personality were much more responsive 
to evidence of conflict and stress than to behavior that indicates per- 
sonal wellness (Honigmann 1954:104; Maslow 1950; Dunn 1959). 
Such personality stress need not have been produced by accultura- 
tion. Acculturation or, to be more exact about the variable that is 
probably crucial, uneven culture change (cf. Mead 1956), un- 
doubtedly encourages personal stress, but stress is also evident in 
American Indian personality as it became known in very early con- 
tact times (Hallowell 1946). Characterological stress continues to 
be found among remote northern people like the Kaska Indians who 
adhere to a way of life not grossly changed from aboriginal times 
(Honigmann 1949). I do not suggest that the American Indians 
have from prehistoric times been subjected to more personality con- 
flict than other people. 

How far anthropologists' accounts of American Indian person- 
ality have been influenced by factors such as reservation life remains 
a question worth investigating in detail. The restricted range of 
occupations, atmosphere of paternalism, and social arrangements 
that relieve the Indian of considerable responsibility for creatively 
solving his problems undoubtedly help to standardize behavior in 
adults and children. Behavior, overt and covert, that is so stand- 
ardized is what the anthropologist observes. Hence, reservation life 
must have personal repercussions, but its influence need not be 
predominantly pathological. My own experience also leads me to 
believe that the transformation of the American Indian personality 
has been less pervasive than superficial evidence of assimilation (in 
clothing, housing, jobs, language, and other elements of reservation 
life) leads some people to believe. Iroquois Indians in New York 
State, Cherokee in North Carolina, Sioux in North Dakota, and 
Makah in Washington do not structure experience precisely like 
their Euro-American neighbors even though they may in some cases 
speak the same language. Anthropology, of course, devotes itself 
to more than the study of only highly exotic cultures or personali- 
ties. However, to an extent truly exotic data are valuable and even 
essential in order to accumulate comparative material on which to 
base universal generalizations. Exotic material can still be secured 
in our continent by someone capable of close, clinical observation 
that dives below the superficial veneer of Americanization. One or 
two summers of field research are insufficient to discover the social 


personality of a community. Culture and personality research re- 
quires an intensive understanding of individuals who must be seen 
over long periods in their environment. With them the anthropolo- 
gist must develop intensive rapport. 

Culture and personality studies began among North American 
Indians with the collection of personal documents, a category in 
which I include autobiographies, biographies, and psychological 
analyses such as Gregorio, The Hand-Trembler (Leighton and 
Leighton 1949) or Devereux's (1951) account of a psychoanalysis. 
American Indians have provided some notable personal documents, 
including Radin's (1920) account of Crashing Thunder, Dyk's 
story of Son of Old Man Hat (1938) , Simmon's (1942) rendering 
of Sun Chief's own life, and Ford's Smoke from Their Fires ( 1941 ) . 
However, the exploitation of this channel to present "person- 
defining" behavior has not been very widely pursued. It is much to 
be regretted that in most cases we do have no more than one first- 
rate life history per culture. Nor have many innovations appeared 
within the life-history approach. Oscar Lewis (1959) is responsi- 
ble for a new departure in his portrayal of five family cultures in 
Mexico, though his approach departs somewhat from the strictly 
personal document. Life histories, as Kluckhohn ( 1945) points out, 
are valuable for the insight they provide into the meaning which 
social forms possess for the members of a given community. They 
are analogous to the case histories which psychiatrists collect from 
patients and study carefully because in those communications the 
patient's style of life is revealed. But anthropologists are ultimately 
interested in more than the record of a specific individual's experi- 
ences. They note Sun Chief's attitudes toward sex not merely as 
one individual's way of handling of a universal situation, but for 
what they tell us about how that aspect of Hopi culture is generally 
experienced — hence, the importance of accumulating personal 
documents from a number of people who occupy different statuses 
in a particular community. 

Culture and personality research has not remained identified 
with life histories. To understand how it came to apply theories 
from child development, psychology, and psychiatry in the study 
of culture, we must note the emergence at the end of the nineteenth 
century of psychology as a science.^ Twentieth century psycholo- 

^ Wundt's Voelkerpsychologie is only indirectly related to the origins of the culture and 
personality movement which, however, shows definite traces of the Volka^eisf School of German 
historians like Ranke and Grimm (Kluback 1956:24). For other antecedents see Meggers 


gists showed increasing interest in the relationship of personaHty 
development (including the breakdown of personality organiza- 
tion) to social conditions (Burt 1957) . Meanwhile, anthropologists 
noted that culture after all is manifested only through individuals. 
This conclusion occurred to Boas, for example, though he did little 
to pursue it. He did, however, transmit his interest to a number of 
his students who were to become extremely influential in the new 
movement (cf. Kluckhohn 1944:596; Mead 1959:14). 

Among those students was Edward Sapir, who, in his paper "Cul- 
ture, Genuine and Spurious" (1924), distinguished between the 
concept of culture as applied to man's whole material and spiritual 
social heritage and to "those general attitudes, views of life, and spe- 
cific manifestations of civilization that give a particular people its 
distinctive place in the world." Sapir was offering a new version of 
an orientation that had long interested certain historians, like those 
of the Volksgeist group in nineteenth century Germany. In subse- 
quent papers (for example, "Cultural Anthropology and Psychia- 
try," 1932) , Sapir advanced the germ of the definition of culture 
and personality which I have offered at the start of this chapter. 
Cultural anthropology, he said, emphasizes the group and its tradi- 
tions but pays little regard to the individuals who make up the group 
and who actualize its traditions in individual variations of behavior. 
Anthropology might focus on persons and see culture in its "true 
locus," namely "in the interactions of specific individuals and, on 
the subjective side, in the world of meanings which each one of 
these individuals may unconsciously abstract for himself from his 
participation in these interactions" — in much the same way as psy- 
chiatry focuses on a whole individual and observes him in his world 
of social relationships. 

If I had to date the actual beginning of culture and personality 
field research conducted in this spirit, I would choose the year 1928, 
the year in which Margaret Mead — a student of Boas — published 
Cofiting of Age in Samoa. However, we are concerned with North 
American Indians. Here the signal event emerged from Ruth Bene- 
dict's (1928, 1932, 1934) preoccupation with characterizing cul- 
tures in psychological terms. In 1934 this brilliant student of Boas 
published Patterns of Culture. The book attempts to characterize 
several cultures in terms of contrasting psychological orientations. 
One chapter of the book, in which she contrasts the Indians of the 
Great Plains with the Pueblo people (Zuni) of the Southwest, will 
illustrate Benedict's approach. 


The Plains way of life reveals a Dionysian quality. In personal ex- 
perience the Plains Indians seek to press beyond the commonplace 
toward excess in order to achieve a certain psychological state. The 
Pueblo Indians in contrast are Apollonian, meaning that they dis- 
trust excess, prefer to keep to the middle of the road, and avoid 
meddling with disruptive psychological states. Benedict saw Plains 
and Pueblo cultures as two configurations. The Dionysian and 
Apollonian emphases reveal themselves in many parts of the con- 
figuration, for example, in response to death. The Plains Indians 
give way to uninhibited grief when a kinsman dies; mourning is 
prolonged, and some people even mutilate their bodies in a form of 
self-torture. The Apollonian Pueblos also react to death with sor- 
row, but people seek to make as little, rather than as much, of the 
event as possible. In each culture area the ideal personality type 
reflects the dominant psychological orientation. The Plains value 
the self-reliant man. By showing initiative in war or hunting, such 
a man achieves honor. The Pueblos have a different ideal. They 
value the mild-mannered and affable man who acts in moderate 
rather than in grandiose or spectacular terms. 

In another chapter of her book, Benedict describes the Kwakiutl 
Indians of the North Pacific Coast of North America. She views 
them not only as Dionysian, but characterizes them as obsessed 
by megalomaniac ideas of grandeur, ideas which express themselves 
in furious competitive feats (potlatches) and in the way chiefs 
seek to gain the best of one another through boasting and mutual 

The inspiration for Benedict's brand of configurationalism came 
not from anthropology, nor from a school of psychology that was 
already current, Gestalt psychology, but from a historian, Oswald 
Spengler (1926) . Note that Benedict's interpretations of cultures 
in psychological terms omits intensive, firsthand study of the peo- 
ple whose behavior she describes. The Plains Indian and Kwakuitl 
ways of life which she characterizes had long vanished and Benedict 
relied on ethnographers' earlier accounts. Culture and personality 
rarely again followed this method but instead put great reliance on 
firsthand field work. For if personality is interpreted solely from 
ethnographic materials which describe a culture, there is danger 
that the actual underlying psychological organization of the people 
who live that culture will be falsified. Explanation will be circular: 
the cultural datum — people behave peaceably and co-operatively — 
will be explained in terms of underlying peaceful, restrained, and 
co-operative motivations. This danger is inherent in Benedict's 


approach, though mostly she avoids f aUing into circularity because 
she does not essay a direct account of personality. She tends to say 
people act as if they had such motives. The safe position is never to 
assume that overt peaceableness or any other cultural trait is moti- 
vated by a similar state, like absence of hostility. It may or may not 
be. The point is that the existence of motives cannot be directly 
inferred from the outward form of behavior. Motivation and cul- 
ture are not isomorphic. Motives must be assessed through studying 
living individuals in depth using clinical methods. Or else the 
various myths, films, and fictions of a community may be inter- 
preted in a clinical manner (see Margaret Lantis' approach de- 
scribed below) . 

How shall Patterns of Culture be evaluated? Some anthropolo- 
gists have condemned the book as subjective and unscientific. In 
some instances such condemnation is motivated by anthropologists' 
unwillingness to admit that their discipline includes a strong hu- 
manistic tradition. Benedict, however, clearly thought of her work 
as scientific. One reason why she may have identified with science 
is that in her day, as in ours, categorizing a piece of research as sci- 
entific surrounds it with greater authority. Patterns of Culticre 
does partake of science, provided we are not too narrow in how we 
define the concept and do not make science identical with the ex- 
perimental testing of hypotheses. Any attempt to generalize knowl- 
edge fits into the scientific tradition. Benedict offered a method for 
generalizing many specific bits of behavior in order to see cultures 
as wholes. 

Patterns of Culture contributed much to stimulate thought con- 
cerning method and interpretation (cf. Nadel 1937; Li An-Che 
1937) . The very fact that people distrusted Benedict's interpreta- 
tions and felt that she was too subjective made them refer back 
to the same evidence she too had used. Her accounts of Pueblo and 
Kwakiutl life have been found to be incomplete. She selected facts 
to draw a picture that would be in accord with the way the Pueblo 
and Kwakiutl themselves ideally view life. She ignored some in- 
stances of behavior that were incongruent with the configuration 
of dominant, ideal interests. Nevertheless Patterns of Cidture re- 
mains timelessly important and in a certain sense indisputably valid 
in the same way that any great interpretation of reality remains 
valid because it expresses fully the aims of its creator. So too a ju- 
dicious historian's work remains viable even after subsequent works 
are written that contain more complete evidence and more up-to- 
date interpretations. 


One paper, written by John Bennett (1946) to review recurring 
disagreements over the interpretation of Zuni and adjacent Pueblo 
cultures, raises methodological implications that go quite beyond 
the field of culture and personality. Bennett examines the interpre- 
tations of Pueblo life made by Benedict and others and sets them in 
opposition to another view of Pueblo culture, one that he calls the 
"repressed" approach. His conclusion is that in each case the values 
of the observer to a certain extent necessarily govern the way he 
structures his data. We cannot determine on empirical grounds 
once and for all which point of view is "right." 

The publications that appeared in the latter part of the decade 
and in the forties reveal that actual field research was already under 
way in the thirties. In 1937 came the first of Landes* reports on the 
Ojibwa (1937, 1938a, 1938b) and Hallowell's (1936, 1937; also 
see 1942, 1946, 1951, 1952) work on another branch of the same 
ethnic group. Hallowell's research among the Ojibwa indicates 
that, although personality development is undoubtedly influenced 
by cultural change, in some respects the personality system is also 
highly autonomous and persists. In eastern North America, his evi- 
dence indicates, the fundamental organization of personality per- 
sisted through two centuries of culture contact. Hallowell's (1946, 
1952) method was to compare the reports of seventeenth and 
eighteenth century missionaries and explorers with the people as 
he knew them. In the early period Europeans characterized the 
Indians as emotionally restrained, stoical, strongly inhibited in the 
expression of aggression, mild in the face of provocation to anger, 
and suppressive of open criticism. In "deeper" or more nuclear 
terms, Hallowell finds in the reports evidence that the aboriginal 
northeastern Indian was anxious lest he fail to maintain the re- 
quired standards of fortitude, express anger and resentment, or pro- 
voke the anger of others. Essentially the same characteristics still 
existed in the relatively unassimilated Ojibwa Indians whom Hallo- 
well observed along the upper banks of the Berens River which 
flows into Lake Winnipeg and even in the more assimilated Ojibwa 
who live farther down the river. Indians who had been in more in- 
tense contact with Euro-Canadians did differ in some respects from 
their more isolated contemporaries. For example, they were more 
extroverted. But the personality core, Hallowell found when he 
scored responses to the Rorschach test given by both Berens River 
groups, was fundamentally the same. No radical psychological shift 
had occurred in the course of acculturation. Later Hallowell moved 


to the still more acculturated Lac du Flambeau Indians in north- 
ern Wisconsin, another branch belonging to the same ethnic group 
of Ojibwa. Here, in spite of heavy culture change and cross breed- 
ing between Indians and whites, he found that the Lac du Flambeau 
people psychologically remained Indians. Obviously these people 
who were being encouraged to live as Euro- Americans would have 
a difficult time adjusting to the demands of their social environment. 
Characterologically they were in another cultural world, says 
Hallowell, anticipating one of the main conclusions of the U.S. In- 
dian Education Research Project which will be described more 
fully below. Presumably the core structure of the aboriginal per- 
sonality was able to resist change because it could get along with the 
traditional characterological system, though this cannot be ac- 
cepted as a full explanation of how that traditional character struc- 
ture manages to be transmitted from one generation to another. 

Certain methodological aspects of Hallowell's work deserve spe- 
cial note. In effect he applied to his three communities a variant of 
the experimental method — actually the only kind of experimental 
method that can be applied in studying living groups of people 
(Chapin 1947) . His procedure involved a fruitful adaption to cul- 
ture and personality research of the method of intercultural com- 
parison, a method which has a long history in anthropology. His 
groups illustrate three levels of acculturation. On Level One were 
the least acculturated, pagan inland Ojibwa of Berens River. Then 
came the Christian lakeside people, among whom aboriginal dwell- 
ings had disappeared along with the old songs and ceremonies. About 
20 per cent of this group were of mixed racial ancestry. On Level 
Three we find the highly acculturated Lac du Flambeau Indians of 
Wisconsin, 80 per cent of whom were racially mixed and all spoke 
some English. The Lac du Flambeau children attended school, their 
families had radios, and in general the people maintained a close 
association with whites. However, at Lac du Flambeau the Mide- 
wewin ceremony had been carried over from precontact times. 

The Rorschach test offered Hallowell a common device which 
he could apply in each group to measure differences in response. He 
tested over 200 people with this instrument, recognizing, of course, 
that it had never been fully validated for cross-cultural use (Hallo- 
well 195 1 ) . One of his findings we have already stated: persistence 
of personality independent of degree of assimilation to Euro-Cana- 
dian or Euro- American culture. Another finding comes from 
counting the number of signs of adjustment that appear in the 


Rorschach responses of each group. Differences in adjustment are 
not significant when the two Berens River communities are com- 
pared to one another, but there is a significant increase in personal 
maladjustment in the records of Lac du Flambeau. For example, 9 
per cent of the Level One records show signs of bad integration 
compared to 18 per cent of the Lac du Flambeau subjects.^ We 
should also recognize Hallowell's lasting interest in social psychi- 
atry, a field he has pursued with the aid of North American cultural 
data. He was one of the earliest anthropologists to distinguish be- 
tween normal and abnormal anxiety ( 1936) . 

Hallowell revealed new possibilities in using the ethnohistorical 
method to reconstruct aboriginal personality. One other example 
of this method may be mentioned, Esther Goldf rank's (1943) work 
on the Teton Dakota. She shows how aspects of Dakota interper- 
sonal behavior — notably aggression — altered in pace with other 
changes in the way of life. Before 1850 the Dakota were horse- 
mounted buffalo hunters and warriors. Ingroup violence was fairly 
common and sprang partly from ingroup rivalries. The rich com- 
peted with displays of wealth. The introduction of liquor by early 
fur traders intensified violence toward the end of this early period. 
Between 1850 and 1 877 increasing contact occurred with the white 
man and there was a growing decimation of the wild buffalo, the 
Indian's mainstay. Aggression was turned outward as wars broke 
out between the Indians and Euro-Americans over the latters' en- 
croachment on the land and on account of broken treaties. When 
the Indians' aggressive energies began to be deflected against ene- 
mies, a need for increased responsibility and in-group co-operation 
arose. It is largely for this reason that ingroup aggression began to 
decline, though competitive displays of wealth by the rich con- 
tinued. What pressures were used to alter personality with respect 
to aggression? The chiefs, whose position had grown stronger, gave 
sermons on the importance of ingroup co-operation. Blood money 
rather than blood revenge was used to settle murder. To borrow 
terms which Anthony F. C. Wallace (1959) has introduced, the 
periodic expression of impulses normally suppressed gave way to an 
emphasis on the lasting suppression of incongruent motives and be- 
havior. A similar phenomenon occurred among the Iroquois after 
their disorganizing contact with Euro-American civilization. For 
a time the Dakota managed to release aggression outward, against 
rival tribes and the United States' troops, but their power to do so 

^For other Ojibwa research see Caudill 1949. and Barnouw 1950. 


was broken following Custer's massacre. Between 1877 and 1885 
the Indian was "crushed." In this third period the buffalo disap- 
peared and the old economy was wrecked. Most of the horses had 
been taken by the victorious army. With the external threat re- 
moved, acute internal aggression again broke forth. The chiefs' in- 
junctions were ignored. But now a strong, foreign, legal system was 
on hand to curb the disruptive trends that had almost free play 
prior to 1850. From 1885 onward the people reluctantly turned to 
making a living as farmers and also to religion. Chiefs entered the 
ministry and became pastors of their people. The Indians eagerly 
adopted one feature of Christianity, the blessedness of giving. Re- 
ligion and law restored ingroup peace and generosity became an 


Goldfrank's work exhibits one difficulty encountered with the 
ethnohistorical method: it too rarely permits psychologically so- 
phisticated inferences of motivation. The nature of the available 
data forces the student to deal largely with the overt features of 
personality or interpersonal relations. 

For most workers the dominant aim in culture and personality 
research has been to throw light on motives and feeling states which 
underlie overt behavior. Applying the Rorschach test in field work 
and interpreting the responses with the aid of Rorschach theory 
constitute one way of reaching the covert area of personality. Of 
course, a person who relies solely on the test and ignores clues to 
covert states present in other instances of overt activity is basing 
his understanding on a very narrow foundation. One cannot infer 
covert phenomena from outward forms without some kind of 
theory, the purpose of which is to specify how to proceed with in- 
terpreting in covert terms what people say, do, make, or write. 
Some form of the psychoanalytic theory (usually not in its most 
extreme, orthodox form) is still the most widely employed adjunct 
to culture and personality research, though the utility of the theory 
for cross-cultural research has been questioned at certain points, 
for example, concerning the universal existence of an Oedipus com- 
plex. However, with regard to defense mechanisms, the impor- 
tance of childhood in personality formation, the overdetermined 
nature of behavior, the motivated nature of dreams, and other 
subjects, psychoanalytic theory has been confidently and on the 
whole successfully utilized. 

I shall not trace the somewhat complicated history of the appli- 

'For other studies of Dakota (Sioux) personality see Erikson 1935), and Macgregor 1946. 


cation of psychoanalytic theory to culture and personality. The 
psychoanalytic approach in anthropology came to maturity with 
the publication of The Individtial and His Society, a book written 
by Abram Kardiner, psychoanalyst, in collaboration with Ralph 
Linton, anthropologist (Kardiner 1939) . The same year saw publi- 
cation of another psychoanalyst's ''Observations on Sioux Educa- 
tion" (Erikson 1939) . TJoe Individual and His Society is not based 
on deliberately organized field work in North America, although 
the authors do briefly examine the Zuni and Kwakiutl Indians and 
also the Eskimo in terms of their theory. The book grew out of a 
seminar jointly conducted at Columbia University by Kardiner 
and Linton. The seminar continued and provided Kardiner with 
material for a second volume. The Psychological Frontiers of So- 
ciety (1945). In this book one American Indian group, the 
Comanche, receives intensive consideration though no fresh data 
were collected for the purpose of this analysis. As a matter of fact, 
the interpretation pertains exclusively to the aboriginal Comanche 
personality, that is, to the period when the Indians were warriors 
and buffalo hunters on the southern plains. 

Since a comprehensive statement of Kardiner's theory is given 
by Thomas Gladwin in another chapter of this book (Chapter 5) 
I need not do so here. We should, however, recognize the emphasis 
which most schools of psychoanalysis put on the early years of life. 
Childhood is the period when the meanings in terms of which indi- 
viduals carry out other aspects of their culture — war, religion, child 
rearing, and many other activities — are established in the personal- 
ity. Ideally, psychoanalytic theory aims to predict the way an adult 
will regard his world and himself in terms of the way he was reared. 
But it is doubtful if an adult social personality can really be pre- 
dicted in this way except in very general and not very useful terms. 
What customarily happens is that the adult covert personality — 
what Kardiner calls the "basic personality type" — is interpreted 
using knowledge of how children are currently being socialized and 
also by drawing simultaneous inferences from adult activity. In- 
stead of really predicting, the researcher attempts to develop a 
plausible explanation which will tie into a neat package both cer- 
tain events of early life and certain selected features revealed by 
adults' overt behavior. 

In 1945 I was enough impressed with the potentialities inherent 
in Kardiner's work and Karen Horney's (1939, 1945) version of 
psychoanalytic theory to apply this approach to the Kaska Indians 


who live in northern British Columbia and southern Yukon Terri- 
tory (Honigmann 1949) . My intentions among the Kaska were to 
identify the emotional qualities which people revealed as they acted 
their cultural roles and account for such qualities in terms of un- 
derlying, dominant motivations. I also hoped to explore the con- 
ditions of early life under which the dominant motivations are 

Kaska social personality is characterized by seven, very much 
interrelated, dominant motivations, or value orientations, each of 
which must be understood in terms of its context and not by other 
definitions which the terms may have. The first of these motives is 
egocentricity, defined here as a high evaluation of personal inde- 
pendence in which interests are self -centered rather than group- 
centered. This motivation colors the way Kaska Indians resist direc- 
tion from sources outside the family. It enters into the positive 
evaluation of work, which guarantees independence and self-suffi- 
ciency in this trapping-hunting economy, and also into the mascu- 
line striving of women, some of whom appear to be in part dis- 
satisfied with their sex role. 

A second dominant motivation is utilitarianism, a concept that 
refers to a practical and resourceful attitude toward the problems 
of living, an interest in concrete rather than abstract thinking. The 
Kaska are present-oriented and little concerned with a remote fu- 
ture. Deference is a third guiding tendency in the nuclear area of 
Kaska social personality. The word denotes an attempt to maintain 
f rictionless human relationships and a concern lest one becomes dis- 
liked and rejected. In conformity with this value orientation, peo- 
ple make requests obliquely, thereby not risking open rejection and 
also not pressing on other people too aggressively. More directly, def- 
erence is expressed by the avoidance of face-to-face quarrels. Hos- 
tility is, however, expressed indirectly and covertly through gossip. 
In other words, hostility is not lacking in Kaska social personality. 
Evidence for it appears in dreams and more overtly in how some 
people act when they are intoxicated, for example, threatening 
others and themselves with violence. The normal suppression of 
interpersonal hostility is very useful for people who live in an 
atomistic social system, one without strong social controls. 

The next dominant motivation, flexibility, is difficult to define 
positively. It denotes a state of mind in which external necessity, 

For a study of the Aymara Indians of Peru whicli employs a similar appro.icli to tlie core 
personality see Tschopik 195 1. 


duty and hurry are subordinated to personal inclination. This state 
reveals itself in an absence of rigidity and in tolerant, even inde- 
cisive, attitudes toward the demands of living. The absence of hurry 
or of rigorous timetables, the people's easy conscience, the non- 
compulsive way in which children are reared and dogs trained, and 
the lack of obsessiveness all express this motive. In certain crisis 
situations flexibility combines with dependence, another dominant 
motivation, to produce procrastination and hesitation. As a result 
of these behaviors, the critical state that confronts the individual 
may grow worse instead of being resolved. The motivating state of 
dependence needs little explanation, though it should be noted that 
this tendency in the character structure is at variance with the em- 
phasis characterologically placed on egocentricity and resourceful- 
ness. It is quite possible for a social personality to reveal inconsistent 
trends which people themselves occasionally have difficulty recon- 
ciling in their day-to-day living. 

Finally there is emotional isolation, perhaps the most dominant 
note in Kaska Indian social personality. The concept includes a 
strong desire to maintain aloofness from emotional experience and 
emotional involvement as well as a tendency to suppress all feeling. 
It is based on a characterological inability to tolerate strong emo- 
tion, including affection. Egocentricity is quite congruent with a 
social organization in which for much of the year families engaged 
in trapping live in relative isolation from one another in the bush 
and under a social system that is without superordinate authorities. 
Sexual constriction is one specific mode in which emotional isola- 
tion is expressed in interpersonal behavior. This form of expression 
shows up in the ambivalence that marks the relations of men and 
women, in the absence of public display of affection between 
couples, in the reluctance to marry (that is, to enter a strong emo- 
tional — even dependent — relationship), and, most dramatically, 
in the behavior accompanying premarital sexual relations. Premari- 
tal sexuality includes considerable preliminary teasing that cul- 
minates in a chase, capture, struggle, and, finally, coitus. Such a 
sequence, I discovered when I lived among the Indians, is often 
difficult to distinguish from actual rape. Girls and also married 
women conceive of the sex act as a hostile encounter, a perception 
they reveal in dreams and in the associations spontaneously given to 
dreams. The promiscuity of adults, since it offers the opportunity 
for sexual satisfaction without risk of emotional involvement, also 
reveals emotional Isolation. 


In general the Kaska world-view wavers between the idea that 
experience is manageable and the idea that life is difficult as well 
as uncertain. The self -view also comprises two conflicting attitudes: 
value placed on self-reliance and a tendency to abandon striving 
and revert to passivity. The former is far more conscious, and much 
more acceptable, than the second. Passivity particularly manifests 
itself in crises, when there is eager reaching out for help (cloaked, 
of course, by virtue of the tendency here called emotional isola- 
tion) and surrender of active striving. 

Emotional isolation is the motivation whose grounding in early 
socialization is easiest to perceive. This value orientation is rooted 
in the way a Kaska mother withdraws emotionally from her child 
when the youngster is between two and three years old. She does 
not outrightly reject the child but spontaneously withdraws show 
of warmth and affection. The mother becomes more impersonal, 
more concerned with herself, or more preoccupied with a younger 
sibling. She shows herself less patient and indulgent to the young- 
ster. In this situation the child unconsciously makes a decision never 
again to invest strong affection in others. The significance of grow- 
ing up and spending all one's life with relatively affectless people 
who serve as role models must not be ignored in understanding how 
the Kaska style of life is acquired. 

Psychoanalytic theory suggests that the striving for independ- 
ence, which also makes up Kaska social personality, is founded on 
the indulgent care of infants. In this highly favorable period of life, 
the Kaska baby develops an unverbalized attitude of confidence in 
himself and hopeful expectations toward the world. These expecta- 
tions are only loosely entrenched, however. They are contradicted 
by the emotional withdrawal that comes as an early shock. The pas- 
sivity of Kaska personality in certain crisis situations can be ex- 
plained as it derives from this traumatic episode and also as it reflects 
the hold which the passive-receptive state of infancy continues to 
exert in the personality. 

The major test of truth that can be applied to this kind of inter- 
pretation is the test of consistency. Is the explanation sufficient, 
reasonable, clear? Does the explanation offered explain the facts 
in noncontradictory fashion? Does the evidence hold together sen- 
sibly? Are contradictions between facts, if they occur, adequately 
accounted for in terms of the theory that is being used? For reasons 
that I shall examine more closely at the conclusion of this chapter, 
anthropologists have become shy of research whose validity can be 


assessed mainly on evidential grounds, that is, by applying tests of 
consistency and reasonableness. 

Culture-and-personality studies directed to the American scene 
thrived in the forties. One development that added enormously to 
our knowledge of American Indians as persons began in 1941 with 
the start of the Indian Education Research Project (also called the 
Indian Personality and Administration project) . This was a co- 
operative venture in which the Committee on Human Develop- 
ment of the University of Chicago was allied with the United States 
Office of Indian Affairs where John Collier was Commissioner 
(Havighurst and Neugarten I955:v-vi; Thompson 1951:12). 
Their general purpose was to examine the whole development of 
Indian children in six American Indian tribes in order to derive 
practical, useful lessons for Indian education. What was happening 
to the personalities of Indians under the impact of American civili- 
zation? An answer to this question, it was believed, would help to 
define the "real needs" and resources of American Indians and 
would serve as a guide for administrators. In other words, although 
the results of the project were expected to contribute substantially 
to general knowledge, the project was designed as action research 
or applied anthropology. Indian Service personnel, mainly teachers, 
nurses, and school administrators, were recruited to do much of 
the field work, but professional anthropologists were also assigned 
to the six groups selected for intensive study. In addition to anthro- 
pologists the project was carried through psychologists, psychia- 
trists, public administrators, linguists, and other specialists. The 
groups for which monographs of findings have been published are 
the Hopi (Thompson and Joseph 1944) , Sioux (Macgregor 1946) , 
Navaho (Leighton and Kluckhohn 1947; Kluckhohn and Leighton 
1946) , and Papago Indians (Joseph, Spicer, and Chesky 1949) . The 
reports on Zia and Zuni Pueblos have unfortunately never ap- 

The approach which these works follow may be called psycho- 
genetic or developmental. With a variety of methodological aids 
(Emotional Response, Moral Ideology, Rorschach, Thematic Ap- 
perception, and other tests) as well as direct observation, the intel- 
lectual and emotional development of children is followed from 
birth to adolescence. The underlying theory draws from psycho- 
analysis, but the various workers are concerned with more than 
the earliest years of life and base interpretations on experiences 
that occur considerably later than feeding, toilet training, or early 
sexual training. 


It is beyond the scope of this chapter to summarize and evaluate 
the whole project. Fair evaluation especially would be difficult, for 
clear evidence concerning how the results of the research entered 
into the administration of the United States' Indians is hard to come 
by. Laura Thompson (1951) has written on the significance of 
the project for which she co-ordinated research activities. Six years 
of field work, she says, were required before a general solution of 
the welfare problem peculiar to each tribe could be formulated. 
The research involved far more than the relationship of personality 
to culture change. Other variables also had to be taken into ac- 
count: ecology, health, social organization, language, arts, crafts, 
ceremonies, and the core values of the people. The main findings 
were, first, that a program of administration which was oriented 
primarily to assimilating the Indians into the general American 
population was highly detrimental to the welfare of Indian com- 
munities and Indian personality. Second, a substantial increase in 
costly schools, health services, and technological aid will not bring 
about rapid assimilation of the Indians into the general population. 
Thompson writes: "We may predict with assurance that the cur- 
rent Indian Bureau pohcy of rapid assimilation and 'liquidation,' 
in so far as it is effectively implemented at the reservation and the 
community levels, will be detrimental to Indian personality de- 
velopment and community welfare." On the other hand, her find- 
ings support the wisdom of the Indian Reorganization Policy which 
had been adopted under the early administration of Commissioner 

Perhaps the best way to give some conception of this research is 
to take a specific tribe and describe findings which are relevant to 
culture and personality there. For this purpose I have selected the 
Hopi Indians (Thompson and Joseph 1944: Thompson 1950). 

The birth of the Hopi child occurs in the mother's home. Shortly 
thereafter rites introduce the newborn individual to his father and 
to the Sun and also initiate a life-long series of gift exchanges be- 
tween the child and his father's clanspeople. The infant spends prac- 
tically all of the first three months of life in a supine position on a 
cradleboard. After this time the cradle is used only as a place to 
sleep until it finally becomes discarded between six months and a 
year. The cradle, it is suggested, probably contributes to the baby's 
feeling of security and also conditions the newborn individual to 
expect restriction. But many other, less physical restrictions will 
appear as the child matures. Weaning comes with little difficulty, 
usually around the age of two years. Cleanliness training is intro- 


duced gradually, without shock. Up until the age of six in boys 
and throughout youth in girls the mother and other females of the 
matrilocal household act as the primary agents of socialization. The 
mother's brother is a source of stricter discipline. The general char- 
acter of early life is permissive but the freedom of the youngster is 
firmly limited in the interests of his physical safety. From such limi- 
tation every Hopi probably gains an early conception of how haz- 
ardous the environment is in his village. Adjustment is more difficult 
for boys than for girls, a generalization that is revealed in boys' 
behavior problems, like thumbsucking, temper tantrums, and steal- 
ing. The explanation lies in the fact that girls grow up in a house 
where they are expected to remain even after marriage. Boys, 
already by the age of four or five, begin to break away from the 
family group and spend more and more time in the kiva (a religious 
structure) or in the fields and on the range. Eventually a young 
man will marry into a strange house and there assume a very mar- 
ginal position. Actually, the boy also gains freedom by breaking 
away from his family around the age of five. In contrast, the girl's 
role remains restricted. She must stay close to home and help her 
mother, and she too experiences conflicts that show up in temper 
tantrums, stealing, and fighting. Psychological tests show that five- 
year olds among the Hopi are more relaxed and spontaneous than 
older Hopi children. For one thing, they are not yet fully disci- 
plined. The girl's inner life at this age is simpler than the boy's; he 
is already quite introverted and shows a pervasive, vague anxiety. 
Initiation into the Kachina cult marks the transition from child- 
hood to youth. The ceremony introduces the child to the Kachinas, 
his ancestors, who send rain and food in exchange for prescribed 
ritual behavior. Initiation means a ceremonial whipping for some 
children, depending on the sodality into which they are initiated. 
Naughty boys, it is said, are usually initiated into the sodality that 
calls for the more severe whippings. The boy is whipped while he 
is stripped naked, but a girl initiate wears her clothes and is beaten 
less severely. Following initiation, public opinion to a considerable 
extent replaces the matrilocal household as the main control over 
the child's behavior. The father remains a source of happiness to 
youngsters but, tests show, the larger community becomes a source 
of fear, punishment, anger, and shame. The tests also reveal the 
child's conception of his family as a source of reward and praise. 
Economic responsibilities also increase for the initiated boy. Both 
sexes restrict their play to the evening. From six to twelve children 


attend day school, an experience that girls particularly welcome be- 
cause it liberates them from the house. At fourteen some boys go on 
to boarding school. As seen in psychological tests, the period from 
eight to ten is a time when outside contacts increase for both sexes. 
Girls are finally aroused from their simple, unquestioning, walled-in 
existence; their imagination develops; their personality becomes 
more complex, more like the boy's. Just before puberty, however, 
boys and girls reveal a tendency to withdraw into themselves, much 
of the earlier spontaneous responsiveness to outside impressions dis- 

The transition to adulthood in Hopi life is not clearcut, though 
marriage marks a profound change in role. Tests probe below the 
surface to reveal what happens in adolescence as the sexual impulse 
rises in consciousness. However vaguely sex is defined by the young 
person, it is not perceived as evil. The force of the sex impulse now 
halts the introversive trends so apparent at the threshhold of 
puberty. An easier acceptance of outside contacts takes place. Boys 
achieve sex indulgence more easily than girls. Hopi girls are not 
allowed to roam around and must avoid showing themselves to be 
boy-crazy. Hence, girls continue to demonstrate more emotional 
withdrawal than boys. 

The Hopi and other Indian samples of children were compared 
to a Midwest, white sample in order to establish differences (Havig- 
hurst and Neugarten 1955). In contrast to the latter, Hopi chil- 
dren derive little happiness from personal achievement. This is 
understandable for they have been taught to avoid any demon- 
stration of achievement. Yet the Hopi youngsters are consciously 
proud of being praised and respected. Tests also show that aggres- 
sion makes Hopi children anxious, perhaps because of the enormous 
pressure that the community exerts against fighting. Work is im- 
portant in their young lives; in how well or poorly he performs it, 
an individual demonstrates whether he is of good or evil character. 
Conscience is reflected through belief in immanent justice — belief 
that morality is sanctioned by an all-knowing unchangeable, and 
unchallengeable external moral power. Belief in immanent justice 
in Hopi children does not decline with age as it does in Midwest 
children. In fact, the belief increases with age! Belief in animism 
decreases more slowly among the Hopi than in the Midwest sample. 

The Rorschach test reveals near-adolescent Hopi children to pos- 
sess a deeply disciplined character structure. These youngsters are 
carefully selective with regard to their emotions; they are cautious 


and restrained. Yet they recognize pleasurable aspects of the world, 
though these must be accorded their due place. Exuberance is toned 
down. The children have average good imagination but it is seldom 
richly fluid, lively, or vivacious. Here again appears the all-per- 
vasive note of restraint. (Compare how closely these findings cor- 
respond to Ruth Benedict's [1934] characterization of the Pueblo 
Indians as Apollonian! Remember, Benedict achieved her insight 
without the benefit of gathering her data clinically.) Instead of be- 
ing primarily concerned with the emotional aspects of impressions 
and events, Hopi children approach the world intellectually and 
imaginatively, though without abandoning themselves to fantasy. 
The Hopi child is cautious, especially in his approach to a new situ- 
ation. He does not become confused by something new. He rather 
firmly accepts or declines what is offered; his behavior sometimes 
makes the Hopi youngster appear stubborn or unshakeable to his 
teachers. The personality reveals a vague, free-floating anxiety 
which is unattached to definite, fear-provoking objects. In this 
character structure we see reflected the "price" that the Hopi child 
pays in order to survive in an environment which he has been taught 
is filled with potential danger and one which for these desert farm- 
ers is actually perilous. The Hopi adapts by limiting his desires, 
emotions, and ambitions. Limitation in turn generates an "inside 
pressure" that lacks any definite outlet. The child feels discomfort 
and fear without understanding that the source of the disturbing 
force is his own overdisciplined self. Such fear is expected in the 
Hopi community and is socially "normal." One area of personality 
remains unaffected by discipline, the area of the instinctual (includ- 
ing sexual) urges — the id. These impulses remain unusually vivid 
and spontaneous. 

Adult Hopi are much given to malicious gossip and frequently 
suspect one another of witchcraft. Such behavior probably origi- 
nates from hostility and anxiety. From whence do hostility and 
anxiety arise? They arise from social relations carried on in a small, 
town-dwelling group, a group that is vulnerable to danger of fam- 
ine and epidemics and whose pressure is a source of anger, shame, 
and punishment. The role of the mother plays a part. As a discipli- 
narian she is a source of anger, shame, and discipline — more to the 
boy than to the girl. Hostility and anxiety are also rooted in the in- 
ability of the child to form deep, emotional attachments with any- 
body, except the mother, a person with whom his relationship is 


Two further developments that brought culture and personality 
to maturity in the forties must be mentioned. First, criticism began 
to be leveled against the new movement. Particularly did critics 
object because, they thought, too much was being claimed for the 
formative years of childhood in the process of personality forma- 
tion. Anthropologists doing such research, themselves deplored the 
excessive weight that, under the inspiration of psychoanalytic 
theory, was sometimes given to early disciplines (Goldfrank 1945; 
Underwood and Honigmann 1947) . But this was only one contro- 
versial feature of the vigorous, new approach. Others too received 
a full, frank, and sometimes hostile airing. In her review of "Re- 
cent Trends in American Ethnology" Betty J. Meggers (1946:186) 
looked with alarm at the way Sapir had been heeded and attention 
was being diverted from cultural to psychological problems. Cen- 
sure, Meggers said, was being met by anthropologists who chose to 
study culture. "That this trend will continue for some time to 
dominate anthropology cannot be doubted," she wrote. "In the 
meantime, however, the province of culture is being neglected." 

Critical notice was not the only indication of the maturity which 
culture and personality had achieved. A second was the appearance 
of two collections of mainly reprinted readings (Haring 1948; 
Kluckhohn and Murray 1948). These, naturally, did not limit 
themselves to data from North America. Both quickly went into 
new editions and were joined by a textbook in culture and person- 
ality (Honigmann 1954) . 

The new decade opened with two contributions from Latin 
America which marked new levels of development. Holmberg's 
(1950) study of the Siriono is essentially ethnographic, but the un- 
derlying problem derives from psychological theory. Where a sparse 
and insecure food supply exists, do frustrations and anxieties cen- 
tering around the hunger drive have major repercussions on be- 
havior? Holmberg found overwhelming evidence for strong anxiety 
responses toward food among the Siriono and he traced their de- 
velopment back to Siriono childhood. In the other work, John Gil- 
lin (1951) examined cultural sources of threat and security 
affecting Indians and Ladinos in a Guatemalan community. 
Rorschach analyses had already appeared comparing these two pop- 
ulations and had also examined the motivational makeup of six 
witch doctors (Billig, Gillin, and Davidson 1947-48) . In the same 
year as Gillin's publication, Oscar Lewis (1951) published his study 
of Tepoztlan, a book important as much for the questions it poses 


pertaining to the re-examination of already studied cultures as for 
itself being a meticulous approach to personality conceived of 
largely as manifesting itself in interpersonal relations. 

One major development of the fifties transcends the North 
American culture area. This is a comparative approach that utilizes 
statistical techniques to test cross-ndhirally the relationship be- 
tween aspects of child rearing (the antecedent variables) and sub- 
sequent personality or cultural variables. The principal work of this 
type is Whiting and Child's (1953) Child Training and Personal- 
ity: A Cross-Ciiltural Study, Several other publications followed in 
the late fifties (Child, Storm, and Veroff 1958; Spiro and D'An- 
drade 1958 ; Barry, Child, and Bacon 1959) . Quantitative inquiries 
and correlation analyses have a long history in cultural anthropol- 
ogy. In this case they have been facilitated by the existence of the 
Human Relations Area Files developed at Yale University. 

The comparative method, relying on statistical tests of relation- 
ship, coincided with a continuing and mounting wave of criticism 
directed against what is still sometimes called the "excesses" of cul- 
ture and personality studies. Orlansky's (1949) literature search 
had already assembled much material showing that, contrary to 
psychoanalytical theory, no consistent or meaningful relationship 
linked early forms of nursing and personality traits in latter child- 
hood. In the next year, further searching questions were asked by 
an anthropologist, psychologist, and two sociologists (Goldman 
1950; Farber 1950; Lindesmith and Strauss 1950) , not to speak of 
Roheim's (1950) strictures directed against members of the "cul- 
turalist school" for rejecting pure Freudian theory as being too 
biological! Against this background let us examine briefly some re- 
cent methodological innovations in North American research, par- 
ticularly those introduced by the Harvard Values Project (Kluck- 
hohn 1951); by Spindler (1952, 1955) in his careful research 
design for studying personality variation as correlated with differ- 
ential assimilation of a foreign culture among the Menomini, and 
by Wallace ( 1952) , who demonstrates how the Rorschach test can 
help in deriving a true modal personality type. 

All research in values is not equally concerned with studying 
personality. For example, Northrop's (1946) and Albert's (1956) 
interests hardly seem to be. But the Flarvard Values Project has 
tended to keep its focus on individuals, and Clyde Kluckhohn 
( 1 9 54 : 69 1 ) said that the work of his colleagues is partly in the field 
of culture and personality. The special attraction of values research 


lies in the fact that it provides a procedure promising a higher meas- 
ure of objective rehabihty than many people would see residing in 
the more subjective approach keynoted in Patterns of Culhire, or 
in the diagnoses of psychoanalytically oriented workers. Projective 
tests it is true also did much to reduce subjectivity and heighten re- 
liability (as far as test protocols, not interpretations, are concerned) 
but they still leave unanswered the question of the validity of the 
test itself. Note that George Spindler (1955), in a work to be re- 
viewed, sharply separates his interpretation of Rorschach data from 
the scored responses which he analyzes statistically. His first and 
main proof of personality differences between categories of people 
is in the objective and statistically defensible scores (pp. 1 22-123) • 
Wallace's modal Tuscarora personality, too, we shall see, is first con- 
structed out of test scores. In using the Rorschach technique, it is 
the validity of the interpretations that presents the problem. 

Vogt's (1951) work among the Navaho can be taken as a fair 
example to illustrate the contribution that the values approach 
makes to understanding people. Some Navaho men who served in 
the U.S. Armed Forces significantly shifted their value orientations 
{cf. Florence Kluckhohn 1950) , for example, dropping the Navaho 
orientation that views man to be subjugated to nature and adopting 
the position that man controls nature. Some veterans also adopted 
a future outlook in place of being primarily oriented to the present. 
All veterans, however, did not assimilate Euro-American values. 
Vogt shows that sociocultural variables, like disruption of the fam- 
ily of orientation as well as the size and structure of that family, 
are conditions which governed the veterans' acculturation. Large 
extended families, to take another specific instance, tended to con- 
serve Navaho values, exerting a negative influence on assimilation. 
The individual's personality adjustment also related to his readiness 
to alter his values. Those Navahos who accepted white values tended 
to be characterized by stronger personal conflicts and insecurity. 

The experimental method that Hallowell pioneered when he ex- 
amined personality and acculturation cross-culturally was ad- 
vanced by Spindler (1955) in his well-designed study of the Me- 
nomini Indians on their Wisconsin reservation. He too relied on 
the Rorschach test. Spindler graded a sample of 68 male Menomini 
Indians (all at least half Indian in ancestry) in terms of degree of 
assimilation. At one extreme of his five-point continuum are the 
native-oriented population, people who obtain subsistence from 
wage work but also continue with hunting and fishing. They con- 


sciously maintain kinship ties and traditional ceremonies. All per- 
sons in this category speak Menomini. In character structure they 
show a passive but not hopeless orientation toward life unmarked 
by strong threat. This narrowly defined personality is hardly suited 
to competitive struggle or to the expression of aggression. They 
keep a damper on emotional expressiveness. 

Next come the peyotists, the members of the peyote ritual group 
who practice a ceremony that is not traditional and in which visions 
are a key feature. They are people over whom the old culture main- 
tains a substantial hold although they do not fully endorse its tra- 
ditions. Characterologically they reveal a quality of hopeless, pas- 
sive soul searching that expresses individual anxiety. 

Then come the reservations' transitional people who have no 
overt ties with the old culture and have adopted a full measure of 
the new way of life. On a deeper level, however, nostalgia for the 
past reveals itself along with identification with Euro-American 
culture. Transitionalists are less passive and more aggressive than 
the native-oriented population. They do not deal with anxiety 
through hopeless soul searching. In them aggression sometimes takes 
explosive forms. 

In fourth place are the lower-ranking, assimilated Indians who 
obtain their living from lumbering and belong to the Catholic 
Church. People on this level of assimilation are no longer passively 
oriented but the character structure is deeply disturbed. 

Most assimilated are the members of the elite-assimilated cate- 
gory, Spindler's fifth category of reservation people. The men hold 
supervisory jobs in lumbering and other fields and also belong to the 
Catholic Church. Personality reveals a quality of ready emotion- 
ality. There are no signs of disturbance as in the previous group and 
little evidence of passivity. 

Spindler's study is notable for several things. It confirms and am- 
plifies the thesis that acculturation has been detrimental for some 
American Indians. Instead of speaking globally of all Menomini 
men, it divides the population into categories based on degree of 
assimilation and demonstrates meaningful psychological differences 
between the categories. This represents a degree of refinement in 
culture and personality research though it does not invalidate gen- 
eralizations based on a community as a whole, generalizations that 
for some purposes continue to be very useful. Principally, Spindler's 
work is meritorious for the precision and objectivity it reveals, 
qualities that it has not been possible to demonstrate adequately 


here in this small summary. But the Menomini study is mainly de- 
scriptive. We miss in it the painstaking examination of socialization 
that would enable us to understand how each of the five categories 
achieves its particular behavior style. The almost complete reliance 
on the Rorschach is unfortunate in one respect. Good clinicians do 
not rely exclusively on one test. There are characteristics of behavior 
that the Rorschach test cannot pick up but which a sensitive ob- 
server could bring out. The loss in objectivity would to my way of 
thinking be balanced by the enriched picture of personality pro- 

Anthony Wallace's (1952) work with the Tuscarora demon- 
strates how the Rorschach test and appropriate statistical pro- 
cedures allow a strictly modal-personality type to be constructed. 
The term "modal personality," of course, was used before Wallace 
but there is a substantial difference between usages. The usual con- 
structs of so-called modal personality (for example, by Honigmann 
1949; DuBois 1944) or of what Kardiner calls "basic personality" 
are really ideal types and not constellations of traits most frequently 
(modally) appearing together in a community (cf. Aberle 1954: 
669). Wallace sampled deliberately and his work deals with true 
modal types. If culture and personality have usually been delineat- 
ing ideal types, then doesn't it follow that certain criticisms of na- 
tional-character studies must be reconsidered? I have in mind par- 
ticularly the criticism which condemns such research as invalid 
when it does not sample a national population by class, region, and 
similar attributes. The heterogeneity of a modern nation need not 
be incompatible with the construction of an ideal personality type 
of the country as a whole, although the usefulness of such a type 
may be queried. 

Bert Kaplan's monograph, A Study of Rorschach Responses in 
Four Ctdtures (1954), is a very astute, experimental appraisal of 
culture and personality method and of the Rorschach test applied 
in such research. In a sense it is a reply to the critics of culture and 
personality research. Kaplan asks whether, objectively'', there is such 
a thing as social personality. Specifically, can the Rorschach test 
pick up personality differences between the culturally different 
communities? Comparing Zuni, Navaho, Mormon, and Spanish- 
American Rorschach records Kaplan proves that, generally speak- 
ing, each of these groups does perform differently in the Rorschach 
test situation and concludes that systematic differences in person- 
ality do occur from one community to another. 


As we come up to the present it is hard to gauge the long-range 
significance of a given piece of work. But the new, good sense that 
Anthony F. C. Wallace (1955, 1956) makes out of certain of the 
ethnohistorical data pertaining to North American Indian accul- 
turation will probably count in future anthropology. He has intro- 
duced the concept of "revitalization" to designate the psycho- 
logical processes that operate in persons during certain kinds of 
nativistic movements. Such movements, he shows, can be inter- 
preted in psychological terms quite meaningfully. The sequence of 
a typical revitalization movement, according to Wallace, begins 
with a period of constantly mounting stress. Over the years people 
look for a way out, for some way to restore a more satisfactory cul- 
ture. Some people "succeed" in effecting rather narrow-base, per- 
sonal "solutions" for their stress through such behaviors as alco- 
holism or neurosis. War and changes in political leadership are also 
tried, and new economic doctrines are advanced, but generally with- 
out much success. At one point a prophetic leader appears. He an- 
nounces a solution that came to him, perhaps from a divine source. 
At this point, assuming that the leader is indeed heeded, revitali- 
zation sets in as order is restored in the community's world of mean- 
ings. People become more satisfied and hopeful; the stressful condi- 
tions of their existence are alleviated, at least for a time. The prophet 
shows an intense concern for cultural reforms. The changes he pre- 
scribes range from minor ritual innovations to institutional rear- 
rangements that add up to a substantially new culture. Wallace 
focuses on the prophet and tries to account dynamically for his 
behavior. Typically prophets have been disturbed people exposed 
to intense personal and social stress. Wallace looks to the level of 
physiological functioning for much of the explanation of the 
prophets' personality resynthesis. When the prophet's stress reaches 
a critical point "the physiochemical milieu for resynthesis is auto- 
matically established." A convulsive effort to redesign his percep- 
tion of the situation occurs and becomes the basis of his teaching. 
Of course, what message the prophet hears and what lines of action 
he recommends cannot be explained physiologically. They depend 
on his prior experience and intelligence. 

Wallace's work is notable for the courageous way in which he 
attempts to fuse social and physiological levels of analysis. He opens 
himself to the charge of being reductionistic, that is, of explaining 
phenomena on one level by phenomena belonging to another system 
of events. But for many people such criticism carries little weight, 


provided that the explanation which is offered really explains what 
is being studied. His explanation, of course, is hypothetical, and we 
may not know for a long time, if ever, whether physiochemical 
changes are indeed associated with prophetic revelations. (For a 
related work of the American Indians' response to the United States 
civilization, see Voget 1956, though he does not write primarily 
from the psychological point of view. ) 

We have said several times that the prevailing tendency in cul- 
ture and personality research is to observe with the aid of appropri- 
ate theory and with clinical exactness living people in their normal 
environment in order to infer the underlying psychological states 
by which they can be characterized. But we have also noted that 
Goldfrank, Hallowell, and Wallace utilized ethnohistorical data 
in pursuing personality studies. The work of Margaret Lantis 
(1953, 1959) demonstrates well how theory can be applied to a 
rich mythology in order to assess personality. Her work with Nuni- 
vak Island Eskimo mythology draws on a close and detailed knowl- 
edge of the people, knowledge based on long-term acquaintance. 
Lantis offers theoretical justification for using myths as evidence 
of psychological processes. Mythology, like folklore, brings out peo- 
ple's objective view of reality and also offers insight into their sub- 
jective perception of what that reality means to them. The sharing 
of myths in a community offers all members an opportunity to 
standardize their views of human behavior and of the rest of the 
natural world. Myths, in other words, constitute an amalgamated 
body of science, philosophy, and religion through which people give 
structure to reality. Radcliffe -Brown (1930-31:63), a British 
anthropologist, in similar terms speaks of a social structure that 
includes not only human society but also that society's relationship 
with its total environment. 

Elimination, sex, intercourse, and other bodily functions are re- 
ferred to very casually in Nunivak Eskimo myths. Their relative 
de-emphasis may, of course, be due to repression, but such an in- 
terpretation is not confirmed by other evidence (for example, 
extant cultural patterns) . Apart from sex, myths indicate that 
the relationship of men and women is quite a complex problem for 
the Eskimo. Men pursue in women an idealized mother image. Yet 
the terms in which the myths portray women (that is, the way men 
perceive them) suggests that men are often disappointed in their 

Nunivak Eskimo individuals seem to possess a firm idea of what 


they want to be and a clear image of the world in which they realis- 
tically strive to attain desired ends. The characters in the myths are 
persistent; usually they are cautious and judicious observers, ra- 
tional beings, willing to admit defeat while at the same time trying 
to overcome it. They are responsible, diligent, and methodical be- 
ings who in most cases prove to be effective in their goal-oriented 
behavior. All these traits, says Lantis, indicate that the Eskimo 
himself has a ''good orientation to reality." Yet, on the ego level of 
functioning the Eskimo personality, judging from the myths, is not 
quite what at first glance it seems to be. The readiness of the people 
in the stories to accede to others' desires and the tendency to be sub- 
missive suggest a restricted ego. Particularly does ego restriction re- 
veal itself in the way the individual in myths is unable to be aggres- 
sive when he has to further his competitive ambition or satisfy 
some other desire. Toward some interpersonal problems the char- 
acters maintain a laissez-faire attitude; they are afraid of impinging 
on others and therefore restrict their own area of assertive activity. 
Close examination of the stories makes it clear that the characters 
obtain objectives not solely by their own efforts but also through 
magic. When a defense is needed against a feeling of inferiority or 
against real ineffectiveness in a tough situation, the people in the 
stories submit to supernatural power. In psychological terms, this 
suggests that the feeling of inadequacy that the Eskimo experi- 
ences in some situations motivates him to objectify his wishes and 
to rely on relatively passive forms of coping. Such a readiness to in- 
hibit vigorous self-assertion may be acquired early in life, Lantis 
suggests, explaining that her evidence for this hunch comes not 
from the myths but from observation of child rearing among the 
Nunivak people. Submissiveness and only the gentlest signs of physi- 
cal assertion suffice to bring the child satisfying rewards. 

We have looked briefly at the id and ego, and now come to ma- 
terial from myths bearing on the superego level of the Eskimo per- 
sonality. A strong superego is evident in phenomena such as repres- 
sion, subconscious compulsion, and other defenses that appear in 
mythology. Furthermore, restraint on a person's physical drives is 
made into an acceptable positive value. Hostility is often expressed 
deviously, that is, by magical means, rather than through direct 
aggression. More clues to superego functioning come from examin- 
ing the many emotional threats that confront the characters. One, 
especially, is significant: being bitten or eaten. Lantis finds an ex- 
planation for this anxiety in the guilt and fear of retaliation that 


Eskimo probably feel for killing and eating the soul-bearing animals 
on which their life depends. Lantis reasons cogently in order to sup- 
port this interpretation: 

. . . these people who are among the world's most effective hunters, that is, among 
the greatest human predators against animals, feel continuous guilt for this very 
effectiveness and so must enter into the myriad small rituals, must observe the 
tabus, load themselves down with amulets, rush to confess what seem trivial 
offenses, practice the magic, in order to reduce their anxiety. . . . The hunter 
must have sensed his own deep hostility against these creatures that so often 
eluded and frustrated him. 

The myths reveal a large stock of defenses that presumably also 
operate in Eskimo personality, including wish fulfillment, avoid- 
ance, denial of reality, projection, rejection, displacement, undoing, 
and others. Yet, in her final assessment, Lantis finds this personality 
not to be a morbid one. Destructive forces in the myths are after 
all combated successfully. The death of a protagonist is rare and so, 
too, are unhappy endings. The myths show "an objective and ef- 
fective people, much too busy meeting the world to think about the 
emotional conflicts within themselves." 

Lantis reports a brief analysis of thirty-two Rorschach records 
from Nunivak Eskimo men and women that at many points cor- 
roborates interpretations derived from the myths. Subjects who 
took the Rorschach are shown to be of "high average" intelligence 
and given to careful, meticulous observation, almost to the point of 
compulsiveness. They reveal high energy, persistence, and extro- 
version. There is a real tendency to conform but no direct evidence 
of submissiveness. The subjects are preoccupied with sex but with- 
out conflict or guilt (preoccupation seems to be concentrated in 
the Rorschach records of adolescents) . The test records reveal signs 
of frustrated aggression, dependence, and oral aggression, for ex- 
ample, revealed by biting and eating) . Repression, too, is shown to 
be a fairly common defense. Lantis's work, unusual for the inten- 
sive exploration which she devotes to a relatively neglected source 
of data is also noteworthy because it is the only full-scale appraisal 
we have of Eskimo social personality.^ 


Practically all culture and personality research in North Amer- 
ica (and, for that matter, in Latin America as well) has been done 

^For relevant materials on other Eskimo see Honigmann and Honigmann 1953 and 1959 and 
Ferguson i960. For quite a different use of folktalks in culture and personality research see Child, 
Storm, and VerofF 1958. 


by antkropologkts from the United States working in their own 
back yard- They have experimented with a variety of frameworks, 
methcxis, and techniques, including the use of autobiography, depth 
interviewing, psychoanalytical formulations, projective tests, and 
the construction of statistical modal types. Anthropology' in gen- 
eral has always encouraged methodological innovation and experi- 
mentatioiL Innovation in culture and personality research has been 
motivated by the desire to do better work. Research workers have 
aimed to secure more objective data, penetrate "deeper" after elu- 
sive material, and by-pa^ the superficial for the presumably richer 
level of unrevealed conscious or unconscious thought. In general, 
anthix^ologists have only exceptionally trusted themselves to make 
the kind of sweeping interpretations that psychiatrists (especially 
those analytically oriented) make with such confidence. Increas- 
ingly, anthropologists doing culture and personaHty research have 
come to resemble the clinical psychologists, w^ho, when they advise 
a psychiatrist, rely closely on their scores and are often dif&dent, 
cautious, and embarrassed as zit =; rhe subjective tenor of their 
diagnosis goes. It is as if their role as interpreters of j>er5onaHty 
conflicts with valu^ they acquired while apprenticing in the ex- 
perimental laboratory. Anthropologists studying personahties have 
also rardy been subjective in the manner of men like De Madariaga 
or Maurois ^rho put great reliance on their intuitive skills and sensi- 

Thirty years of field work gave time to zry many approaches, but 
they have scarcely been sufficient (considering the available man- 
powo") to investigate more than a small fraction of the indigenous 
New World population. A few culture areas are well represented 
in culture and personality Kterature but for many our knowledge 
is spotty jtidffd The Southwest has been weU studied, but all tribes 
have not received the same amount of attention (KJuckhohn 1954: 
689) .People like the Xavaho and Hopi have been repeatedly visited. 
Ihey are our best laboratori« for future problem-oriented re- 
search. Considerably less thoroughly studied in the Southwest are 
groups like the Papago and Apache. California, the Great Basin, 
Plateau, and North Pacific Coast have been sampled out only ex- 
ceptkmally by more than one field worker. No matter how reliable 
his methods may be, no man can go very far in one short season. 
Quite a bit of work has been done on the Plains; enough for Glad- 
w^in (1957) to suggest that the cultural unity of the area may not 
be accompanied by much homogeneity' of basic p>ersonaht}'. He ad- 


mits that he has compared only two typical tribes, the Comanche 
and Cheyenne, and is aware that for the second of these, very limited 
personality data are available. Ethnohistorical data pertaining to 
New York State Iroquois Indians have been intensively utilized 
for research — more perhaps than the surviving Iroquois themselves. 
Several anthropologists have recently been studying personality 
among the North Carolina Cherokee and we should soon know how 
that community fits into the continental picture. (For a synthesis 
that does not, however, incorporate all available material see Gulick 
i960: Ch. 8-9.) The Seminole represent a continuing, viable cul- 
tural enclave, although one that is hard to work with. In the far 
North the situation is striking: a number of excellent Algonkian 
studies (mostly of Ojibwa-Chippewa communities) , one detailed 
Athapaskan monograph, and, apart from Lantis's work, little con- 
cerning the popular Eskimo! For Latin America the total picture 
is far more spotty. 

Just as culture areas have been spottily covered and with vary- 
ing degrees of intensity so methodology has been divergent from 
one group to another. If we are really to compare the Navaho, Hopi, 
and Ojibwa, don't we have to do among the Hopi and Ojibwa what 
has been done among the Navaho and apply to the Navaho some 
of the questions asked in the other groups? Against this suggestion 
runs the preference to approach each new piece of work with a 
fresh mind (Mead and Wolfenstein 1955:5) . The whole issue may 
revolve around personal inclination. Why not restudy the Berens 
River Ojibwa using the life-history or Kardiner's psychogenetic ap- 
proach? How about a thorough study of the Kwakiutl using Rors- 
chachs? This brings up the value of revisits, preferably by different 
anthropologists, to communities that were studied some time ago. 
Sixteen years had passed since the Kaska Indians were studied. It 
would be appropriate to discover what has happened on the covert 
and overt levels of Kaska personality and in Kaska culture. Indica- 
tions are that tremendous theoretical advances will come in anthro- 
pology when research workers who possess different methods, or 
at any rate a healthy skepticism concerning some of their pred- 
ecessors findings, systematically re-examine the dozens of in- 
tensively studied communities of the world. 

Having complained about spotty coverage in North America, 
let us admit that we know enough to begin to develop wider gen- 
eralizations and comparisons (cf. Kluckhohn 1954:693). A num- 
ber of reports, for example, suggest quite convincingly that a high 


degree of psychological homogeneity characterizes the American 
Indian. Recently a portion of the available data was assembled in an 
admittedly undocumented form by George D. and Louise S. Spind- 
ler ( 1957) . The psychological features which they discovered to be 
most widely exhibited among Indians are: "nondemonstrative emo- 
tionality and reserve" accompanied by a high degree of control over 
in-group aggression; autonomy of the individual; ability to stoi- 
cally endure deprivation and frustration; high value on bravery; "a 
generalized fear of the world as dangerous" a proclivity for practi- 
cal joking; "attention to the concrete realities of the present" (in 
Rorschach argot, the large D approach) , and dependence on super- 
natural power that one strives purposefully to obtain. The picture 
of homogeneity is even more clear cut if we limit ourselves to north- 
ern forest people, Algonkians and Athapaskans. Emotional re- 
straint, for example, appears to be a highly reliable characterization 
of these Indians. Other common traits include a high value placed 
on deference in interpersonal relationships, personal resourceful- 
ness, and individualism. People do not attempt to tell others what 
to do. Authoritarian attitudes and leadership behavior are sup- 

Another line of constructive synthesis for which we are ready is 
to relate particular personality syndromes to technology, social 
structure, and other segments of culture. Hallowell and I have sug- 
gested a relationship between the relatively atomistic social systems 
of northern hunters and their personality. He also perceives con- 
sistency between the inhibition of overt aggression and use of 
sorcery. Laura Thompson (1948) relates Indian world-views to 
bases of subsistence. In the hunting world-view, man conceives of 
himself as a helpless supplicant for power on which he depends for 
success. It comes to him from a universal power pool through dis- 
parate nonhuman entities, chiefly animals, whom he obtains as per- 
sonal guardians. This world-view persists even among agricultural- 
ists in North America but there it is altered. Where people develop 
a more systematic control of the food supply, they no longer con- 
ceive of themselves as helpless supplicants of power which derives 
from disparate power sources. They become power entities in their 
own right and the power source also becomes more clearly struc- 

Assessment of culture and personality research in any area of the 
world can scarcely fail to note the plethora of theoretical problems 
which have been generated by culture and personality research (cf. 


Inkeles and Levinson 1954). The discussion which follows in part 
reflects thinking that developed while work was being done with 
North American personality materials, but it also applies to work 
done in other areas of the world. 

For example, there is the question of how child rearing leads to 
the formation of adult personality configurations. Not that any- 
body doubts the learned nature of personality or would any longer 
ignore the significance of the later years for socialization. But what 
is learned in early childhood? Before the child can verbalize, how 
can we know what cognitive and emotional learning occurs? How 
does early, basic learning continue to influence later learning and 
direct the individual's world and self views? (The theory of cog- 
nitive dissonance, while it doesn't say wholly new things, speaks 
systematically and might fruitfully be applied to the process of 
personality development.) The accumulated materials on person- 
ality from North American Indians and other areas of the world 
are sufficient for at least beginning to develop an anthropologically 
satisfactory theory of socialization. 

How certain core areas of personality are able to persist despite 
change in other areas of culture is a theoretical problem directly 
instigated by research conducted with North American Indians. 
To what extent is such persistence bound up with socialization, 
language, or mode of ecological adaptation? The solution to the 
problem may well lie in an imaginative theory such as Friedl's 
(1956) designed for the Chippewa (Ojibwa). Incessant change 
was characteristic of the aboriginal culture. It has continued with 
culture contact and supports the persistence of personality. 

In noting possibilities for research in North American culture 
and personality — people to be visited or revistied and generaliza- 
tions to be drawn — we must also assess whether the flow of man- 
power is adequate for this research. Anthropology in this country 
does not want for serious graduate students and creative minds. But 
are they turning to problems of culture and personality in propor- 
tion to their growing number? I have noted diminishing enthusi- 
asm for culture and personality research since the thirties and 
forties.*' In the balance of this paper, I shall examine some reasons 


° The editor of the American Anthropologist (Vol. 6i, p. 498) reports that 47 manuscripts 
falling into the category of culture and personality were submitted (not all were published) 
between 1955 and 1958 or 10 per cent of the total (498). Social organization was in top 
position (10 1 manuscripts, 20 per cent) and then came ethnology — ethnography, method-theory, 
and acculturation with 82, 8t, and 57 articles each (the percentages are 17, 16 and 11 re- 
spectively). The criteria used in classification are not given. In a survey that I recently did 


for this withdrawal of interest and also attempt to resolve some of 
the methodological problems that may be discouraging students 
from entering this field. 

Part of the reason for the lack of support of culture and per- 
sonality research is given in these words of Nadel (1957:189) : 

The advance of any science is punctuated as much by the disappearance of old 
problems as by the emergence of new ones. This is little better than a truism if 
we have in mind problems disappearing and discussions or controversies ceasing 
because the issues in question have been resolved. But often it is not a question of 
solution; rather it is a question of changes of viewpoint and interest. The old 
problems are abandoned because they no longer seem important; the controversies 
cease because all that can be said has been said; and if certain questions still re- 
main unanswered, they are yet shelved in spite of it, or perhaps because of it — 
because one realizes that they are unanswerable and should be replaced by other, 
more profitable, ones. 

The change of interest came when new problems opened up in 
adjacent areas of the discipline, particularly with regard to social 
structure and linguistics. These new problems attracted graduate 
students faced with choosing thesis topics as well as full-fledged 
professionals. But this explanation makes us want to know what 
caused culture and personality to lose appeal. Why couldn't it meet 

Several things succeeded in promoting dissatisfaction with cul- 
ture and personality. Instead of proving a challenge, the barrage of 
criticism released in the forties and early fifties proved to be a deter- 
rent. Why did it have a deterring reaction? The answer lies in the 
growing climate of empiricism and operationalism, the high evalu- 
ation of objectivity, and the stress put on objective reliability. The 
positivist conception of science which had long captivated anthro- 
pology and had become the dominant intellectual force in American 
academic life was incompatible with certain aspects of the new ap- 
proach (cf. Kroeber 1915, 1935, 1936) . Foundation support could 
best be commanded by establishing that one's problems were amen- 
able to treatment by procedures generally accepted to be scientific. 
The notion that anthropology is a humanity as well as a social sci- 
ence has been lost (Honigmann 1959b) . If, as is generally assumed, 
scientific method is a unitary thing, then anthropology must con- 
form as closely as possible to the methods used in those disciplines 
that were indisputably in the scientific tradition as currently con- 
fer a biennial review, I came across many papers that took a psychological view of cultural 
phenomena (Honigmann 1959a). But many of those papers hardly represent what I would call 
culture and personality research and are not by anthropologists. 


ceived. To the extent that culture and personaUty could not be re- 
directed along new lines, it lost ground. 

Recently I listened to a discussion concerning two variant inter- 
pretations of the same data from an American Indian community. 
The anthropologists agreed on the facts, but they disagreed when 
it came to ascertaining their psychological meaning for the Indians. 
For one thing, the researchers probably did not really know the 
people very well and hence were handicapped for interpreting 
their data. They also lacked a sufficiently powerful theory in which 
they believed enough to apply it to their facts. But more pertinent 
is the question they faced of proving any one interpretation to be 
objectively more true than the other. How could any reconciliation 
between interpretations be verified empirically? This is an unhappy 
state of affairs for men to contemplate who wish to model them- 
selves after campus colleagues who follow more rigorous methods. 
(Note that this particular difficulty would not have arisen had the 
psychologically minded ethnologists retained faith in one theory, say 
psychoanalysis. Their deductions would have been guided by psy- 
choanalytical principles. Logical reasoning would have brought 
back someone who went beyond the basic postulates of the theory. 
The fact that the insights obtained by psychoanalytical formula- 
tions could not be checked operationally would also not have been 
unduly distressing. But, very likely, back in the twenties and 
thirties it was these very characteristics of psychoanalytical psy- 
chology that made anthropologists decide against following Freud 

One might properly argue that somebody who really wishes to 
study personality as it develops and functions in one set of cultural 
conditions or another doesn't care what his work is called — whether 
science, history, or art. Furthermore, according to some philoso- 
phers, no hard and fast line separates science from other modes of 
understanding (Polanyi 1958). Everybody agrees that experimen- 
tation is not the essence of science. Nor is the central criterion even 
prediction — what can the paleontologist predict? The field worker 
in anthropology is mainly concerned with communicating his un- 
derstanding of the way of life he researches. His work, then, should 
be appraised by how meaningful is the understanding which it offers 
and what it contributes to the wider understanding of man. I have 
long thought that novels are among the most perceptive means of 
gaining insight into ways of life that a skillful or sensitive writer 
authentically grasps. 


Rival interpretations of personality by men who really know 
their people might receive the same kind of attention that is ac- 
corded to rival views of events in history. The final resolution of 
the dispute would have to wait until fresh data are accumulated, 
new field work is undertaken, or a better theory comes to hand. I 
am convinced that we need more perceptive studies of persons whose 
behavior is standardized in different fashions. To obtain such in- 
formation, we need sensitive students willing to immerse themselves 
thoroughly in exotic ways of life and, by whatever means recom- 
mend themselves, come to know the covert and overt sides of the 
people they study. The care, thoroughness, authenticity, level of 
interpretation, and the underlying degree of understanding which 
such studies will achieve will greatly vary from one case to another, 
but they should not be judged by standards foreign to the problem 
in hand. 


Aberlf, D. F. 

1954 Comments on Southwestern studies of culture and personality, by 
Clyde Kluckhohn. American Anthropologist 56:697-700. 

Albert, E. M. 

1956 The classification of values: a method and illustration. American 
Anthropologist 58:221-248. 

Barnouw, V. 

1950 Acculturation and personality among the Wisconsin Chippewa. Mem- 
oirs of the American Anthropological Association No. 72. 
Barry, H., Ill, I. L. Child, and M. K. Bacon 

1959 Relation of child training to subsistence economy. American Anthro- 
pologist 61:5 1—63. 

Benedict, R. 

1928 Psychological types in the cultures of the southwest. Proceedings of the 

International Congress of Americanists 23:572-581. 
1932 Configurations of culture in North America. American Anthropologist 

1934 Patterns of culture. Boston, Houghton Mifflin Co. 

Bennett, J. 

1946 The interpretation of Pueblo culture: a question of values. Southwest- 
ern Journal of Anthropology 2:361-374. 

BiLLiG, O., J. GiLLiN, and W. Davidson 

1947-48 Aspects of personality and culture in a Guatemalan community: ethno- 
logical and Rorschach approaches. Journal of PersonaUty 16:153-187, 


Burt, Cyril 

1957 The impact of psychology upon education. In The yearbook of educa- 
tion 1957. London, Evans Brothers Ltd. 

Caudill, W. a. 

1949 Psychological characteristics of acculturated Wisconsin Ojibwa chil- 
dren. American Anthropologist 5 1 :409— 427. 

Chapin, F. S. 

1947 Experimental designs in sociological research. New York, Harper & 
Child, L L., T. Storm, and J. Veroff 

1958 Achievement themes in folk tales related to sociaUzation practice. In 
Motives in fantasy, action, and society, J. W. Atkinson, ed. Princeton, 
D. Van Nostrand Co., Inc. 

Devereux, G. 

195 I Reality and dream: psychotherapy of a Plains Indian. New York, Inter- 
national Universities Press. 
Du Bois, C. 

1944 The people of Alor. Minneapolis, University of Minnesota Press. 
Dunn, H. L. 

1959 What high-level wellness means. Canadian Journal of Public Health 

Dyk, W. 

1938 Son of Old Man Hat. New York, Harcourt Brace & Co. 
Erikson, E. H. 

1939 Observations on Sioux education. Journal of Psychology 7:101—156. 
Farber, M. L. 

1950 The problem of national character: a methodological analysis. Journal 
of Psychology 30:307—316. 

Ferguson, F. 

i960 Eskimo personality in the light of nine Rorschachs from Great Whale 
River Eskimo. Research Previews (Institute for Research in Social 
Science, University of North Carolina) 8, No. 1:8-13. 
Ford, C. S. 

194 1 Smoke from their fires. New Haven, Yale University Press. 
Friedl, E. 

1956 Persistence in Chippewa culture and personality. American Anthro- 
pologist 58:814-825. 


1951 The culture of security in San Carlos. Middle American Research In- 
stitute, Publication 16. 

Gladwin, T. 

1957 Personality structure in the Plains. Anthropological Quarterly 30:111— 


1943 Historic change and social character: a study of the Teton Dakota. 
American Anthropologist 45:67-83. 


1945 Socialization, personality, and the structure of the Pueblo society (with 
particular reference to Hopi and Zuni). American Anthropologist 47: 

Goldman, I. 

1950 Psychiatric interpretation of Russian history: a reply to Geoffrey Gorer. 
The American Slavic and East European Review 9:151-161. 


i960 Cherokees at the crossroads. Chapel Hill, Institute for Research in 
Social Science, University of North Carolina. 

Hallowell, a. I. 

1936 Psychic stresses and cultural patterns. American Journal of Psychiatry. 

1937 Temporal orientation in western civilization and in a preliterate society. 
American Anthropologist 39:647-670. 

1942 Acculturation processes and personality changes as indicated by the 
Rorschach technique. Rorschach Research Exchange 6:42—50. 

1946 Some psychological characteristics of the northeastern Indiana. In Man 
in northeastern North America, F. Johnson, ed. Papers of the Robert S. 
Peabody Foundation for Archaeology, 3 . 

195 1 The use of projective techniques in the study of the socio-psychological 
aspects of acculturation. Journal of Projective Techniques 15:27-44. 

1952 Ojibwa personality and acculturation. In Acculturation in the Amer- 
icas, proceedings and selected papers of the XXIXth International Con- 
gress of Americanists, Sol Tax, ed. Chicago, University of Chicago 

Haring, D. G. (ed.) 

1948 Personal character and cultural milieu. Syracuse, University of Syracuse 

Havighurst, R. J. and B. L. Neugarten 

1955 American Indian and white children: a sociopsychological investiga- 
tion. Chicago, University of Chicago Press. 

Holmberg, a. R. 

1950 Nomads of the long bow. The Siriono of Eastern Boliva. Smithsonian 
Institution, Institute of Social Anthropology, Publication No. 10. 


1949 Culture and ethos of Kaska society. Yale University Publications in 
Anthropology, No. 40. 

1954 Culture and personality. New York, Harper & Bros. 

1959a Psychocultural studies. In Biennial Review of Anthropology 1959, B. 

Siegel, ed. Stanford, Stanford University Press. 
1959b The world of man. New York, Harper & Bros. 


1959 Notes on Great Whale River ethos. Anthropologica n.s. i : ro6-i 21. 


1939 New ways in psychoanalysis. New York, "W. W. Norton & Co., Inc. 
1945 Our inner conflicts. New York, "W. W. Norton & Co., Inc. 


Inkfxes, a. and D. J. Levinson 

1954 National character: the study of modal personality and sociocultural 
systems. In Handbook of social psychology, Gardner Lindzey, ed. Two 
volumes. Cambridge, Mass., Addison-Wesley Publishing Co. 

Joseph, A., R. B. Spicer, and J. Chesky 

1949 The desert people: a study of the Papago Indians of southern Arizona. 
Chicago, University of Chicago Press. 

Kaplan, B. 

1954 A study of Rorschach responses in four cultures. Papers of the Peabody 
Museum of American Archaeology and Ethnology, Vol. 42, No. 2. 
Kardiner, a. 

1939 The individual and his society. New York, Columbia University Press. 
1945 The psychological frontiers of society. New York, Columbia University 

Kluback, W. 

1956 William Dilthey's philosophy of history. New York, Columbia Univer- 
sity Press. 

Kluckhohn, C. 

1944 The influence of psychiatry on anthropology in America during the 
last 100 years. In One hundred years of American psychiatry, J. K. Hall, 
G. Zilboorg, H. A. Bunker, eds. New York, Columbia University Press. 

1945 The personal document In anthropological science. In The use of per- 
sonal documents In history, anthropology and sociology, L. Gottschalk, 
C. Kkickhohn, and R. Angell, eds. Social Science Research Council 
Bulletin 53. 

195 1 A comparative study of values in five cultures, hi Navaho veterans: a 
study of changing values, by E. Z. Vogt. Papers of the Peabody Museum 
of American Archaeology and Ethnology, Vol. 41, No. i. 

1954 Southwestern studies of culture and personality. American Anthro- 
pologist 56:685—697. 

Kluckhohn, C. and D. Leighton 

1946 The Navaho. Cambridge, Harvard University Press. 

Kluckhohn, C. and H. A. Murray (eds.) 

1948 Personality in nature, society, and culture. New York, Henry A. Knopf. 

Kluckhohn, F. 

1950 Dominant and substitute profiles of cultural orientations. Social Forces 

Kroeber, a. L. 

19 1 5 Eighteen professions. American Anthropologist 17:283-288. 

1935 History and science in anthropology. American Anthropologist 37: 

1936 So-called social science. Journal of Social Philosophy 1:317-340. 

Landes, R. 

1937 The personality of the Ojibwa. Character and Personality 6:51-60. 
1938a The abnormal among the Ojibwa Indians. Journal of Abnormal and 

Social Psychology 33:14-33. 


1938b The Ojibwa woman. Columbia University Publications in Anthro- 
pology, No. 31. 
Lantis, M. 

1953 Nunivak Eskimo personality as revealed in the mythology. Anthro- 
pological papers of the University of Alaska 2:109-174. 
1959 Alaskan Eskimo cultural values. Polar Notes 1:35-48. 
Leighton, A. H. and D. C. Leighton 

1949 Gregorio, The hand-trembler: a psychobiological personality study of a 
Navaho Indian. Papers of the Peabody Museum of American Archae- 
ology and Ethnology, Vol. 40, No. i. 

Leighton, D. C. and C. Kluckhohn 

1947 Children of the people. Cambridge, Harvard University Press. 
Lewis, Oscar 

195 1 Life in a Mexican village: Tepoztlan restudied. Urbana, University of 

Illinois Press. 
1959 Five families: Mexican case studies in the culture of poverty. New 
York, Basic Books, Inc. 

Li An-Che 

1937 Zuni: some observations and queries. American Anthropologist 39: 
LiNDESMiTH, A. R. and A. L. Strauss 

1950 Critique of culture-personality writings. American Sociological Review 

Macgregor, G. 

1946 Warriors without weapons. Chicago, University of Chicago Press. 
Maslow, a. H. 

1950 Self -actualizing people: a study of psychological health. Personality 
Symposium 1:11—34. 
Mead, M. 

1928 Coming of age in Samoa. New York, William Morrow & Co. 

1956 New lives for old. New York, William Morrow & Co. 

1959 An anthropologist at work: writings of Ruth Benedict. Boston, Hough- 
ton Mifflin, 
Mead, M. and M. Wolfenstein (eds.) 

1955 Childhood in contemporary cultures. Chicago, University of Chicago 
Meggers, Betty J. 

1946 Recent trends in American ethnology. American Anthropologist 48: 
Nadel, S. F. 

1937 The typological approach to culture. Character and Personality 5:267- 

1957 Malinowski on magic and religion. In Man and culture, Raymond Firth, 
ed. London, Routledge & Kegan Paul. 

Northrop, F. S. C. 

1946 The meeting of east and west. New York, The Macmillan Co. 


Orlansky, H. 

1949 Infant care and personality. Psychological Bulletin 46:1—48. 
PoLANYi, Michael 

1958 Personal knowledge: towards a post-critical philosophy. Chicago, Uni- 
versity of Chicago Press. 

Radcliffe-Brown, a. R. 

1930—31 The social organization of Australian tribes. Oceania 1:34—63, 206—246, 
322-341, 426-456. 

Radin, p. 

1920 The autobiography of a Winnebago Indian. University of California 
Publications in American Archaeology and Ethnology 16:381—473. 


1950 Psychoanalysis and anthropology. New York, International Univer- 
sities Press. 

Sapir, E. 

1924 Culture, genuine and spurious. American Journal of Sociology 29: 

1932 Cultural anthropology and psychiatry. Journal of Abnormal and 

Social Psychology 27:229—242. 

Simmons, L. (ed.) 

1942 Sun Chief, the autobiography of a Hopi Indian. New Haven, Yale 
University Press. 
Spengler, O. 

1926 The decline of the west. 2 vols. New York: Alfred A. Knopf. 

Spindler, G. D. 

1952 Personality and peyotism in Menomini Indian acculturation. Psychiatry 

15:15 I— 160. 
1955 Sociocultural and psychological processes in Menomini acculturation. 
University of California Publications in Culture and Society, No. 5. 

Spindler, G. D. and L. S. Spindler 

1957 American Indian personality types and their sociocultural roots. Annals 
of the American Academy of Political and Social Science 311: 147—157. 

Spiro, Melford and R. G. D'Andrade 

1958 A cross-cultural study of some supernatural beliefs. American Anthro- 
pologist 60:456—466. 

Thompson, L. 

1948 Attitudes and acculturation. American Anthropologist 50:200—215. 
1950 Culture in crisis. New York, Harper & Bros. 

195 I Personality and government. Mexico, D. F., Ediciones del Instituto In- 
digenista Interamericano. 

Thompson, L. and A. Joseph 

1944 The Hopi way. Chicago, University of Chicago Press. 

Tschopik, H., Jr. 

195 I The Aymara of Chucuito, Peru. i. Magic. Anthropological Papers of 
the American Museum of Natural History, vol. 44, pt. 2. 


Underwood, F. and I. Honigmann 

1947 A comparison of socialization and personality in two simple societies. 
American Anthropologist 49:557—577. 

1956 The American Indian in transition: reformation and accommodation. 
American Anthropologist 58:249-263. 
VOGT, E. 2. 

195 1 Navaho veterans: a study of changing values. Papers of the Peabody 
Museum of American Archaeology and Ethnology, Vol. 41, No. i. 

Wallace, A. F. C. 

1952 The modal personality structure of the Tuscarora Indians, as revealed 
by the Rorschach test. Bureau of American Ethnology, Bulletin 150. 

1955 Stress, personality change, and cultural creativity. Paper read at meeting 
of the American Anthropological Association, Nov. 17, 1955. Mimeo- 

1956 Revitalization movements: some theoretical considerations for their 
comparative study. American Anthropologist 58:264-281. 

1959 The institutionaUzation of cathartic and control strategies in Iroquois 
rehgious psychotherapy. In Culture and mental health, M. K. Opler, ed. 
New York, The Macmillan Co. 
Whiting, J. W. M. and I. L. Child 

1953 Child training and personality: a cross-cultural study. New Haven, 
Yale University Press. 


chapter 5 



National Institute of Meiital Health 


Much of Oceania is comprised of islands, islands which are charac- 
teristically fairly small, tropical, and separated from adjacent lands 
by open ocean, sometimes by vast stretches of ocean. Obvious ex- 
ceptions to this generalization are the great land mass of Australia 
and the large islands of New Guinea and New Zealand. The pre- 
ponderance of smallish tropical islands has inevitably created both 
limitations and challenges to the pursuit of anthropological studies 
in the area. 

The most severe limitation is set by the thinness of the archaeologi- 
cal record, at least as it has been revealed thus far. In part at least 
this must be ascribed to the high rates of oxidation and biotic decay 
characteristic of warm climates, heavy rainfall, and proximity to 
the sea. Even artifacts tough enough to survive such conditions are 
likely to find the ground washed away beneath them. Although 
some recent archaeological work has been more encouraging, 
Oceania is far from having the solid foundation of prehistory found 
elsewhere in the world. Added to this is a short and very fragmen- 
tary historical record. The result is a focus primarily on the here 
and now, on the present characteristics of populations and cultures 
rather than on their antecedents. 

Granting an inadequate or nonexistent developmental perspec- 
tive, the islands of the Pacific frequently provide the challenge of a 
nearly ideal research setting. The physical anthropologist can find 
a relatively stable, isolated, and homogeneous breeding population 
on which to base his studies. Cultural homogeneity within an island 
can bring similar clarity to the study of social structure and cul- 
tural dynamics. Furthermore, the small size and isolation of many 



island communities permit the detailed description of the totality of 
a finite population. Within a setting of this sort it is often possible 
to define and examine all of the interpersonal and intergroup rela- 
tionships which determine the relevant social environment of an in- 

Finally, if one disregards the large land masses of New Zealand, 
Australia, and New Guinea, the ecology and basic economy of the 
smaller islands of Polynesia, Micronesia, and Melanesia have many 
common, almost uniform, attributes. They are tropical, with abun- 
dant rainfall at least part of the year, and are favored with trade 
winds. In all three areas there are islands which are clustered to- 
gether, and some which are widely scattered and isolated. There are 
flat, sandy coral islands on atolls ("low" islands) and steeper, 
usually larger, volcanic ("high") islands. On all of them the soil is 
relatively poor, favoring principal reliance on root and tree crops, 
and discouraging domestication of animals for meat. This leaves the 
ocean as a primary source of protein. Metal is generally lacking. 
This is only a partial list, which could be extended to include tech- 
nology, health conditions, transportation, etc. On this common 
ecological base one finds a wide range of social and political organi- 
zations, value systems, and personality types. A special opportunity 
thus exists for comparisons between one group and another with a 
number of variables fairly well controlled. 

A particularly fruitful comparison might be made with respect 
to the response of these various island peoples to foreign, especially 
European, contact. The circumstances of this contact were again 
rather uniform. After the early explorers came traders and mis- 
sionaries, and in many areas whaling ships from New England seek- 
ing provisions and release from shipboard life. The traders were fol- 
lowed by more stable commercial arrangements, especially the 
exploitation of coconut and other crops through foreign-operated 
plantations, or through resident traders. Despite the relative simi- 
larity of this experience, the response to it appears to have been 
markedly different. Here the differences can be mentioned in only 
a general and impressionistic way. 

The Micronesians, with few exceptions, retained their core cul- 
ture — especially their social organization, values, and economy — 
while adopting a wide array of superficial technological changes. In 
Polynesia the changes were more sweeping, often devastating. This 
was especially true on the larger island groups such as Hawaii, Ta- 
hiti, or the Marquesas. Even on the smaller atolls, foreigners — mis- 


sionaries, traders, or administrators — were granted more leadership, 
and therefore more opportunity to effect pervasive changes in val- 
ues, in political structure, and in other ways. In Melanesia there 
was frequently hostility, suspicion, and bloodshed, with minimal 
acceptance of foreign leadership except when imposed by force. 
Yet Melanesia has also experienced the sweeping fantasy of embrac- 
ing foreign culture, or at least material culture, in the cargo cults. 
These are bizarre outbursts in which a whole population may, for 
example, destroy its possessions and await a ship full of foreign 
goods — a ship which of course never comes. 

Even though these characterizations are obviously overgeneral- 
ized, it is clear that there were striking differences in response. There 
were of course special factors. Polynesian girls looked especially 
attractive to American men. Melanesia was a source of slave labor 
for "blackbirders." And so on. But these factors do not obscure the 
fact that the people themselves responded differently. Personality 
differences must have played a major role. Conversely, the psycho- 
logical impact of these changes must have varied widely. Oceania is 
therefore an unusually inviting area in which to make systematic, 
comparative studies of culture change, including its psychological 
dimensions. Culture change, however, is only one of many ways in 
which the special character of this area lends itself to research, and 
particularly to research in culture and personality. 

In evaluating the work that has been done to date, we must bear 
these conditions in mind. It is thus particularly legitimate to ask 
what real contributions to our understanding of personality (as 
well as of culture) have emerged from Oceania. After reviewing 
what has been done, I will return to this question in the assessment. 

The review of the literature which follows will be concerned with 
noting the landmarks. It is not in any sense an encyclopaedic inven- 
tory of all work done thus far. Nor does the supporting bibliog- 
raphy pretend to exhaust the literature.^ In particular, I will not 
dwell upon those studies which, while developing data potentially 
useful for the elucidation of personality dynamics, have not been 
developed in this way either by the authors or by others. The mono- 
graph of the Berndts ( 195 1 ) on sexual behavior in Western Arn- 

^I have been greatly aided by a quite complete bibliography through 1954 prepared by the 
University of Hawaii Pacific Islands Studies Committee (Vinacke ei al. 1955). To the serious 
student of the area this three-part bibliography can be invaluable. Taylor's (195 1) Pacific 
bibliography is a standard reference; at the present writing an expanded and updated edition is in 
preparation. The Journal de la Societe des Oceanisfes, published by the Musce de I'Homme in 
Paris, provides an annual bibliographic review of Oceania. 


hem Land or Warner's Black Civilization (1937) would be cases 
in point. 

Geographically and culturally I include work done in Australia, 
New Guinea, and the broad expanses of Melanesia, Polynesia, and 
Micronesia. Among these, Micronesia is a late entrant. During the 
formative years of culture and personality research, Micronesia was 
under the exclusive political control of Japan. The few Japanese 
anthropologists who worked in these islands were seemingly not 
interested in this new field of study, so the first culture and per- 
sonality studies began in Micronesia only after the Second World 

Obviously, inclusion in this survey is based on the locus of field 
work, not on the nationality of the researcher. With the exception 
of Ernest Beaglehole of Victoria University and his students, and 
possibly of Stanley Porteus of the University of Hawaii, none of the 
researchers whom I discuss are residents of the area under consider- 
ation. Reo Fortune is a New Zealander, but no longer lives there. 

Chronological Review and Evaluation 

Oceania can claim a twenty-year beat on the rest of the world. 
It was the locus of the first systematic field work among non-Euro- 
pean peoples designed to enrich the interpretation of ethnology by 
the insights of psychology. This comprised one of the explicitly 
stated aims of A. C. Haddon in organizing the Cambridge Anthro- 
pological Expedition to the Torres Straits. The published report 
(Myers and McDougall 1903) deals almost entirely with the sen- 
sory modalities and would not now be included within the purvey 
of culture and personality as the field has evolved. But its undeniable 
historical significance rests on the fact that Haddon sought, as 
collaborators for Rivers (who was himself trained in psychophysi- 
ology) and Seligman, two psychologists who were felt to be com- 
petent in the most fruitful procedures of the scientific psychology 
of the day. However, at the time the Torres Straits Expedition was 
in the field in 1898, Sigmund Freud was at work on the manuscript 
of The Interpretation of Dreams. It is an historical fact, but not 
necessarily a stroke of undiluted good fortune, that the study of 
culture and personality has come to depend almost exclusively upon 
the line of inquiry being initiated sixty years ago by Freud rather 
than that envisioned by Haddon. 

Certainly considerable credit for determining this trend is due to 
Geza Roheim, the most orthodox and loyal of Freudian psycho- 


analysts to concern himself with non-European personality. As 
early as 1925 he published a book entitled Atisfralian Toteinisvi: 
a Vsycho-analyticd Study in Anthropology. Some years earlier, in 
191 3, Freud had completed his first major work on religion, Totem 
and Taboo. This drew heavily upon the secondhand ethnographic 
data assembled by Frazer, Robertson Smith, and others, especially 
as these pertained to Australia. Among other things, Freud assumed 
that Oedipal conflicts were shared by all peoples. These conflicts 
had their beginnings in primitive family groups wherein at inter- 
vals the sons banded together to kill and sacrifically eat their father. 
The tabooed totem animal survives as a substitute for the father. 
In the literature of the day both totemism and primitiveness were 
emphasized as characteristic of the Australian aborigines, hence 
Roheim's early interest in Australia. After publishing the book 
noted above, which was based on published accounts, Roheim 
set out to see for himself. He undertook field studies in several parts 
of the world (Roheim, 1932) , but spent the longest time with the 
Aranda of Central Australia. He explored at first hand the symbolic 
residue of the primal feast and the conflicts arising from living in a 
primitive horde dominated by an older male. Roheim deserves credit 
for being willing to test his beliefs under rugged field conditions, not 
a fashionable pastime in his day. Furthermore, because Malinowski 
had denied the existence of Oedipal conflicts in matrilineal societies 
(see below) , Roheim also went to matrilineal Normanby Island near 
Malinowski's Trobriands, gathering evidence to refute Malinowski 
{cf. Roheim 1950). For all his enthusiasm, Roheim's work is no 
longer cited with any frequency by anthropologists. He was willing 
to go so far in attributing symbolic and historical significance to 
cultural acts that many anthropologists find it difficult to take his 
work seriously, and therefore tend to dismiss his conclusions (e.g., 
Lessa 1956) . But the lack of continuing attention to Roheim's pub- 
lished writings is deceptive. His early work was widely read and 
initiated or expanded an interest in psychoanalysis in many of the 
influential pioneers in the field of culture and personality. Among 
others, Edward Sapir, Margaret Mead, and Clyde Kluckhohn have 
acknowledged Roheim's important impact on their thinking. 
Roheim's contribution to the field, then, is paradoxical: practically 
no one accepts his conclusions, but their stimulating effect was 
nevertheless very great. 

Roheim also undoubtedly provided part of the impetus for the 
writing of the first monograph which undertook systematically to 


examine a hypothesis of personahty dynamics in the Hght of sohd 
ethnographic data, and in accordance with acceptable anthropo- 
logical standards of interpretation. This was Bronislaw Malinow- 
ski's Sex and Repression in Savage Society ( 1927) . Malinowski drew 
upon his extensive data from the Trobriand Islands to re-examine 
some aspects of Totem and Taboo. As noted above, he rejected the 
universality of the father-son Oedipal conflict, essentially on the 
ground that, in the Trobriands at least, discipline is in the hands of 
the mother's brother and not the father. He also discussed Freud's 
more anthropologically acceptable development of the psychologi- 
cal and social dynamics which support exogamy and the incest ta- 
boo. Roheim was undoubtedly justified in criticizing Malinowski 
for an incomplete understanding of Freud's writings. Malinowski 
was unfortunately the first of many anthropologists who have over 
the years criticized psychoanalytic theory on the basis of an inade- 
quate and watered-down understanding of its implications (r/. La 
Barre 1958) . But Malinowski's excursion into psychoanalytic the- 
ory nonetheless established a precedent for anthropologists. It was, 
in effect, the first anthropologically "respectable" substantive study 
in culture and personality. 

Malinowski's encyclopaedic and highly literate ethnographic 
accounts of the Trobriander Islanders (1922, 1929, 1935) have also 
made important contributions to culture and personality through 
their use by others in developing new lines of analysis. As we shall 
note later, Kardiner started with these materials (as summarized by 
Du Bois) in developing his particular approach (Kardiner 1939) . 
More recently Dorothy Lee (1950) made a major contribution to 
cognitive theory and psycholinguistics in her paper on "Lineal and 
Nonlineal Codifications of Reality," based entirely on Malinowski's 
published accounts of the Trobriands. 

Meanwhile, Margaret Mead had gone to Samoa and in 1928 pub- 
lished on the basis of this field work the first of a series of studies 
of personality development and integration in Oceania. This was 
followed by a comparable monograph on Manus in 1930, and then 
in 1935 by a book describing and comparing personality in three 
contrasting New Guinea cultures — Arapesh, Mundugumor, and 
Tchambuli (the three monographs appeared together in Mead 
T 9 3 9 ) . Accompanying her books on personality, in most cases, were 
solid ethnographic monographs in the best anthropological tradi- 
tion. Concurrently, the impact of Mead's approach was clearly evi- 
dent in Reo Fortune's study of Dobu (1932). Bateson's (1958) 


book centered upon the naven ceremonies of the latmul, first pub- 
Kshed in 1936, marked the beginning of a long collaboration with 
Margaret Mead. Especially to be noted is the large-scale 1936-39 
team study of Bali (especially Bateson and Mead 1942, and Mead 
and Macgregor 195 1 ; a full bibliography is to be found in Mead and 
Wolfenstein 1955:95-98). The many publications resulting from 
the Balinese research reflect a deliberate attempt to develop more 
effective techniques and more rigorous methodologies in support of 
the lines of inquiry established in Margaret Mead's earlier work. 
Aspects of this research which can only be mentioned here, but 
which deserve careful examination, range from systematic exploi- 
tation of photographic techniques, through detailed studies of mu- 
sic, dance, ritual, and drama, to theoretical analysis of social 
equilibrium (Bateson 1949) . Finally, there is the account of Manus 
upon her return there in 1953 (Mead 1954b, 1956) when she found 
and described an extraordinarily successful cultural transformation 
and reintegration. This transformation, as analysed by Theodore 
Schwartz (Mead and Schwartz i960), comprises a valuable addi- 
tion to the literature on messianic movements; they found a com- 
plex interplay between a long-term transformation movement and 
a short-term cargo-type cult. These publications, of course, com- 
prise only a small fraction of the contributions of Margaret Mead 
and her colleagues, and are concerned only with work in the Pacific 
area. But they perhaps define the substantive core of her work and 
methodological influence. The point of view reflected throughout 
her research is effectively (and often charmingly) synthesized in 
Male and Female (1949). 

A more systematic statement of Mead's methodological premises 
is to be found in her retrospective evaluation of the national charac- 
ter studies undertaken during World War II (Mead 1953) . Mar- 
garet Mead was one of the leaders in this challenging attempt to 
construct, at a distance, a basis for predicting the behavior of peoples 
of foreign nations (Mead and Metraux 1953). Anthropologists 
have always been notably reluctant to make predictions, and now 
they were asked to do so by extrapolation from a few informants 
and such documentary materials as could be collected. Mead based 
her approach upon the conception of the individual in his culture 
developed throughout her work and Bateson's in the Pacific: "Any 
member of a group, provided that his position is properly specified, 
is a perfect sample of the group-wide pattern on which he is acting 
as an informant." (Mead 1953:648) The discussion, and sometimes 


controversy, surrounding this methodological approach served to 
make explicit many of the assumptions hitherto lying below the 
surface of stated culture and personality theory. 

Since it would obviously be impossible to summarize separately 
even the few works listed above, it is useful to attempt a general 
evaluation of her work to provide a basis for comparison with that 
of others. Special aspects of some of these studies will be considered 
later in this chapter. 

In evaluating Margaret Mead's work, one fact scarcely needs 
underlining: she is a pioneer. From the outset she did field work in 
Oceania explicitly directed toward the understanding of the varie- 
ties of human personality and the mode of their development (cf. 
Mead 1959). She returned to write monographs of major and lasting 
value years before any other anthropologist (excepting her own 
colleagues) undertook a comparable task. True, Malinowski's field 
work was done during the first World War, and Linton was in the 
Marquesas in 1920-22. But the psychological implications of their 
data were only elucidated as afterthoughts — useful and intelligent, 
but still afterthoughts — many years subsequent to leaving the field. ^ 

A striking example of Mead's pioneering receives too little atten- 
tion, especially in view of the current surge of interest among social 
scientists in the work of Piaget. A primary purpose of Mead's 1928- 
29 field work in Manus (Mead 1932) was to examine the assump- 
tions of Piaget (and of Levy-Bruhl) that the less "logical" (by 
European standards) thought of children was a function of their 
immaturity, and that the thought processes of primitive people 
were analogous to those of children in our society. Using a variety 
of ingenious psychological measures, she found that Manus children 
actually analysed situations in a far more matter-of-fact ("logi- 
cal") fashion than characterized the animistic reasoning of their 

^ Margaret Mead is also an effective and dedicated crusader in the cause of bringing anthro- 
pological insights to bear on the problems of our society. She has translated and focused 
anthropological material upon education, mental health, child development, technical assist- 
ance, and a variety of other fields. There are scores of journals in other professional and popular 
fields to which she has been the first anthropological contributor. Furthermore, her contribution 
has frequently had a clearly discernable effect on the thinking in that profession. Bridging the 
gap between anthropology and a variety of other fields of endeavor often requires a daring leap, 
a leap which some anthropologists feel frequently ends with an agonizing wrench. Without 
laboring this point, I will only suggest that when Margaret Mead is, for example, talking to 
educators she is concerned with improving and enriching our schools, not with meeting the 
canons of anthropological rigor. Her contributions to the anthropological literature provide a 
quite ample basis for judgment of her work as an anthropologist, and I will confine myself to 
these. But I offer my personal cheers to a person willing to balance research with an equal 
commitment to translating the insights so derived into the language and problems of any ac- 
tivity concerned with helping mankind. 


elders. She offered several possible explanations and then arrived at 
the well documented conclusion that, Piaget and Levy-Bruhl to the 
contrary, "Animistic thought cannot be explained in terms of in- 
tellectual immaturity." 

Mead's work in culture and personality rests upon the same con- 
ceptual underpinnings of psychodynamics which are common to 
other workers in the field. Her interest in the possibilities of using 
this body of theory in anthropological research was first stimulated 
in the early 1920's by the writings of contemporary psychoanalysts. 
As her own theoretical position was developing she worked inten- 
sively, among others, with Erik Erikson, Lawrence K. Frank, Kurt 
Lewin, John Dollard, and Edward Sapir. Her approach is perhaps 
most differentiated by the biological and social matrix within which 
she sees these forces operating. She is concerned with the biological 
endowment and biological changes which shape a person's being, 
with the total social environment which surrounds the growing 
child and the adult, with how the child perceives and interprets this 
environment, with how the environment is interpreted, explicitly 
and implicitly, to the child, and with those figures in the social en- 
vironment who are the agents of interpretation and learning. The 
structure of a particular society not only channels all relationships 
and activities within it, but also determines the manner in which an 
individual lives and learns his life. In this approach she hews more 
closely than many to a view of socialization and personality devel- 
opment as a process of enculturation, of the gradual learning of the 
integrated totality of attitudes and feelings and behaviors which 
comprise the culture. Inherent in this is a concern with the process 
and nature of learning, and with the consistency between the pat- 
terns of experience in a variety of learning situations (Mead 1953 ) . 
Related also to this is the intriguing methodological exercise of Mead 
and Macgregor ( 195 1 ) in their photographic analysis of the learn- 
ing by Balinese children of a single facet of behavior, patterns of 
motor activity. In addition to its methodological emphasis, this 
monograph examines the psychoanalytic concept of body zones. 
Analytic theory posits successive concern of the individual with 
the oral, anal, and genital zones. Mead and Macgregor accept this 
formulation for our culture. The Balinese, however, have a greater 
focus on the total body, and in particular on the visual and tactile 
stimulus of the skin. This question is thus raised of how other cul- 
tures shape the interpretation by each individual of his own bio- 
logically given body. 


In this monograph Mead also succinctly characterizes her ap- 
proach to the study of socialization: "Cultural analysis of the child- 
rearing process consists in an attempt to identify those sequences 
in child-other behavior which carry the greatest communication 
weight and so are crucial for the development of each culturally 
regular character structure" (Mead and Macgregor 1951:27). 
Communication is a two-way process, involving interaction in both 
directions. Mead believes the interactive nature of socialization rep- 
resents one of the major new concepts which anthropology brought 
to the study of personality: 

From the cultural anthropologist has come the recognition that cultural forms 
emerge from other cultural forms. Stated genetically, this means that parents and 
children are a continuously interactive system, not a one-way system in which the 
child (impelled upward by a set of specific drives) simply meets a series of ob- 
stacles (in the form of institutions) that, if it is sufficiently mutilated by them, 
it will then proceed to alter. (Mead and Metraux 1953:39) 

Mead's view of socialization studies, her own as well as others', is 
well summarized in her chapter in the Manual of Child Psychology 
(Mead 1954a) . 

Inherent also in Mead's approach is the premise that learning is 
continuous. A person who is growing old, for example, changes in 
behavior not only because of the physiological changes taking place 
within him, but also because he learns to behave in the way the cul- 
ture expects old people to behave. She shares the common focus on 
childhood as the time in which the major dimensions of personality 
are established, but her scheme equally permits substantial changes 
in these constellations through the years which follow. 

Mead's analytic and descriptive procedure has sometimes been 
referred to as "configurational," thereby implicitly identifying her 
work with that of Ruth Benedict. Benedict and Mead were close 
collaborators, and Benedict often followed an approach very close 
to that of Mead, as in her notable paper on "Continuities and Dis- 
continuities in Cultural Conditioning" (Benedict 1938) . But Mead 
is not typically concerned with delineating the broad themes of a 
culture as exemplified in Benedict's P^/Zer//^ 0/ CzJ/7/r^ (1934) , the 
epitome of configurationalism. Mead does point up the consistency 
in feeling tone and attitude from one nexus of interpersonal be- 
havior to another, where such consistency is discernable, but she 
makes no assumption that a common theme must necessarily be 
sought in all important arenas of action in a culture. Benedict's ap- 
proach is global and open-ended. Mead, on the other hand, con- 


stantly reverts back to the social system and structure, and to the 
biological determinants of behavior. 

The distinctive significance of Margaret Mead's work, outside of 
its pioneering nature and the substantive contributions to knowl- 
edge it represents, can best be considered after reviewing the ap- 
proaches developed by later workers in the culture and personality 

Another "first" in the study of personality falls in the still largely 
neglected area of intelligence and cognition. In 1929 Stanley D. 
Porteus, a clinical psychologist trained in Australia but with most 
of his professional career in the United States, undertook field work 
with the Arunta in Central Australia, and more limited work in 
Northwest Australia (Porteus 193 i). He administered, primarily 
to children, a variety of intelligence and performance tests, prin- 
cipal among these being his own Maze Test. This test, as its name im- 
plies, consists of a series of mazes on paper which the subject is asked 
to trace. Porteus later did comparable work among the Bushmen 
of the Kalahari Desert in South Africa (Porteus 1937), and also 
utilized the test extensively in obtaining comparative data among 
the various ethnic groups available to him at the University of 
Hawaii. Of all the tests of mental ability thus far generally available 
for use in cross-cultural settings, there is some reason to feel that 
the Porteus Maze is the "fairest" in the sense that it appears to be the 
least strange and confusing to non-Europeans (r/. Masland, Sara- 
son, and Gladwin 1958:271-72) . 

As a result of this work Porteus was able to formulate a formida- 
ble list of cautions to be observed by anyone attempting cross- 
cultural intelligence measurement, cautions which, despite later 
elaboration by Klineberg and others, were more often than not 
ignored in the years to follow. He also devoted considerable thought 
to the nature of mental ability and its measurement. He concluded 
that any test designed to measure the kinds of mental ability valued 
in our culture would fail to tap those intellectual resources which 
would be useful to a person in another culture where the approach 
to thinking and problem solving might take different directions. 
This applies as much to the Maze as to any other test. However, he 
also made an important but usually overlooked distinction with 
respect to the purposes of measurement: if one is concerned with 
making comparisons between the essential intelligence of two 
groups in an absolute sense, a test built around the concepts of think- 
ing taught in one culture cannot be used validly in another. But if 


one is interested in identifying those persons in another culture with 
the greatest potential for being trained to think our way, as for ex- 
ample in recruiting for schooling, a test devised for our culture can 
be valid. Furthermore, in the latter context, it becomes crucial to 
minimize the degree of strangeness which the test and the testing 
situation evoke among people who are unfamiliar both with the ma- 
terials used and with the whole idea of a test. The Porteus Maze per- 
haps best meets this latter criterion. It would appear useful for 
anthropologists interested in cognitive development to explore 
whether this is really so, and if it is to develop further the poten- 
tialities of this approach. Thus far, however, the Maze has received 
only passing attention from anthropologists. 

As noted earlier, two monographs appeared in the early 1930's 
which reflected Margaret Mead's influence, but which were also of 
major importance in their own right. One was Reo Fortune's Sor- 
cerers of Dobu ( 1932) . Ddbu is a Melanesian island near Malinow- 
ski's Trobriands. The society is so riven with hostility and suspicion 
that Ruth Benedict labeled it "paranoid." Fortune was trained as a 
psychologist and strongly influenced by Freud and W. H. R. Rivers, 
especially the latter's Conflict: and Dream ( 1923 ) . Before going to 
Dobu, Fortune himself had published a book on dream interpreta- 
tion. The Mind in Sleep ( 1927) . In spite of this background, For- 
tune did not systematically address himself to personality as such, 
although he did pay consistent attention to psychologically relevant 
aspects of Dobuan culture. His psychodynamic orientation, how- 
ever, made his work highly appropriate for use by Ruth Benedict in 
counterpoint to Zuni and Kwakiutl in Patterns of Ctdture. It is, in 
fact, in the latter context that Dobu is probably most widely known. 

The other monograph is Gregory Bateson's Naven, recently re- 
published with additional theoretical discussion (1958). Bateson 
did not attempt a full descriptive ethnography, concentrating in- 
stead on exploring the implications of several ceremonies, especially 
the naven ceremony, among the latmul of New Guinea. His analysis 
of these ceremonies led to the formulation of two important new 
constructs. One of these, eidos, will be discussed later in this chapter. 
The other is schizmogeuesis. Schizmogenesis describes those forces 
in society which are centrifugal, that is, which increase the social 
distance between individuals or groups. Schizmogenesis can be 
complementary, as in dominance-submission relationships, or sym- 
metrical, as in rivalry. The centrifugal effect of schizmogenesis de- 
rives both from social dynamics and from personality, and both 


these factors are culturally determined. Marriage in latmul, for 
example, is socially defined as a dominance-submission relationship 
(complementary schizmogenesis) . It also brings together two peo- 
ple with culturally defined different male and female personalities. 
Yet marriages persist (sometimes) in spite of the forces which tend 
to drive the couple apart. However, whereas in latmul schizmogenic 
forces are strong enough to make any equilibrium precarious, a 
comparable analysis of Bah (Bateson 1949) revealed that stabilizing 
forces are so effective that schizmogenic sequences can never get 
started. Bateson therefore became interested in these counterforces 
which keep centrifugal tendencies from going to the extreme of 
destroying all social relationships. This led him to seek controlling 
stabilizing mechanisms which would return the social system to 
balance. In collaboration with a number of persons in various fields 
he turned to theories of mechanics, physics, and mathematics con- 
cerned with feedback and other mechanisms responsible for main- 
taining systems in a steady state of dynamic equilibrium. This field 
of inquiry is now referred to as cybernetics. Collaboration with per- 
sons in fields so exotic to anthropology seems to have created a lack 
of communication between Bateson's thinking and that of all but 
a handful of anthropologists. If this is true, it is a matter we should 
view with alarm. Anthropologists who are content merely to feed 
their cultural data into equations provided for them ready-made by 
personality psychologists can remain union members in good stand- 
ing, but why should a person who reaches out to develop radically 
new equations of human behavior move beyond the pale of anthro- 
pological discourse? 

In the later 1930's Ernest Beaglehole began a series of field re- 
searches in Oceania which he and his students at Victoria Univer- 
sity are continuing to the present. Although trained in his native 
New Zealand and in London primarily as a psychologist, Beaglehole 
studied with Sapir and others at Yale in 193 1-34 and became a well- 
qualified anthropologist. He has consistently provided general cul- 
tural data quite as full as that adduced by other anthropologists 
working in culture and personality. His first field work was accom- 
plished prior to his return to the Pacific, with the Hopi of Second 
Mesa (Beaglehole and Beaglehole 1935). In 1936 he returned to the 
Pacific, first to the University of Hawaii and then to Wellington. 
He initiated a series of ethnographic and culture and personality 
studies in Polynesian societies, including Pukapuka (Beaglehole and 
Beaglehole 1938, 1941), native Hawaiians (Beaglehole 1939), 


Tonga (Beaglehole 1940, 1941), Maori in New Zealand (Beaglehole 
and Beaglehole 1946), and Rarotonga and Aitutaki (Beaglehole 
1957) . In addition he has published a number of important theo- 
retical papers in culture and personality. He has directed a group of 
his students in a large-scale interdisciplinary study of personality 
development in Rakau, a Maori community in somewhat different 
circumstances than Kowhai, the New Zealand community studied 
by the Beagleholes themselves. The Rakau study is notable for its 
extensive experimentation with various methodological approaches 
to the use of projective tests (Beaglehole and Ritchie 1958) . Five 
monographs, in addition to several short papers, have appeared on 
the Rakau research thus far: James E. Ritchie 1956, Mulligan 1957, 
Jane Ritchie 1957, Earle 1958, and Williams i960. 

Beaglehole's contribution to culture and personality in Polynesia 
would be notable alone for the sheer quantity of solid, insightful re- 
search he has contributed to the literature. He has, in addition, made 
a number of clarifying theoretical observations, especially his 1944 
paper on "Character Structure." Here he considered the cultural 
directives governing interpersonal behavior, and their relationship 
to individual personality and behavior deviation, observing that 
when a person 

... is acting according to the major directives he is really acting according to 
a personal organization or structure of his own needs, emotions and thoughts 
which is in congruence with the emphases of the major directives themselves. In 
other words the person has developed a character structure in response to the 
specific pressures of his own culture. When a person acts idiographically, he is de- 
termined by a personal variant on this character structure, that is, by the specific 
drives of unique personality. A person's integrations can be predicted when it is 
known that his personality corresponds rather exactly to the character structure 
of the group. One is often at a loss to predict the course of a person's integrations 
when how different or how alike his personality is to this character structure is 
not known, (p. 148) 

This position is in many respects similar to that of Margaret Mead; 
in each case attention is directed to the shaping of personality by the 
totality of expectations and pressures exerted on and communicated 
to a person by other persons sharing the same culture. Explanatory 
concepts must then emphasize the conditions under which be- 
haviors, attitudes, and feelings are learned by living with others who 
already share such attributes. This is in contrast to explanations 
which lay stress on individual emotional reactions to a succession of 
experiences shared with others in childhood. 

Beaglehole has also gone considerably deeper than most psychol- 


ogists or anthropologists into questions of cognitive structure raised 
by the administration of intelHgence and other tests to non-Euro- 
pean people. The following discussion of his findings on Aitutaki 
bears on this. 

In the cross-cultural measurement of intellectual capacity the psychologists' 
skill and techniques do not yet appear to be adequate to measure differences in 
quantitative amounts of latent intelligence. But test results are still valuable in 
so far as they can be used to indicate the existence of cross-cultural qualitative 
differences in intellectual or cognitive organization. Two aspects of Aitutaki cog- 
nitive organization seem to be suggested by the present results. The first concerns 
the fact that the culture itself does not place value on problem-solving. In its 
technological aspect Aitutaki culture is extremely simple. Results are achieved 
by the simple application of rules traditionally inherited. This is not to say that 
judgment is not required of the successful fisherman or cultivator, but the num- 
ber of variables within his control are so few that complicated judgments are 
hardly ever required. Success in farming and fishing or even in many aspects of 
social life is more likely to be achieved by the application of rules learned by rote, 
rather than by the use of principles applied by reason. Cognitive organization, 
therefore, is likely to be rather simple in structure and largely formed by exper- 
ience derived through the rote learning of repeated lessons. (1957:221) 

The second characteristic aspect of Aitutaki thinking is the fact that it func- 
tions mainly at a perceptual, rarely at an abstract level, and at a perceptual level 
which may be significantly different from the perceptual level thinking of the 
Western European. . . . The way perceptual relations are noticed will be a func- 
tion of a given culture. How the relations, once noticed, will be abstracted and 
generalized about will also depend on the interests and training available in the 
culture concerned. The children of Aitutaki have plenty of experience of coloured 
objects or variously shaped objects, but their culture teaches them to be interested 
mainly in the objects and not in their abstracted shapes, colours and patterns. 
Therefore the quality of their thinking will reflect this perceptual orientation, 
and imaginative thinking either of a controlled or a free fantasy type will be rare. 
This quality of Aitutaki thought again receives confirmation from the limited 
use of imagination in Rorschach records. (1957:22 2-2 23) 

We shall return to the discussion of cognitive process and problem 
solving in the final portion of this chapter. For the present it will 
suffice to note that whereas Porteus and others went no farther than 
to note factors in the tests which interfere with the performance of 
non-Europeans, Beaglehole's discussion goes beyond this to consider 
the differences in learning and thinking which actually create dif- 
ferences in performance. He is also concerned with how these are 
related to the demands of the culture. I would myself raise a further 
question, whether the concept of "learning rules by rote" does not 
itself imply more European-type verbalization of the learning proc- 
ess than in fact obtains. At very least, the descriptive label "rote 
learning" is almost certainly an oversimplification. 


In 1939 a book was published which had a large share in crystal- 
lizing anthropological thought on the relationship between per- 
sonality and culture, and setting the pattern of research for much 
of the present generation of students of the field. This was Abram 
Kardiner's The Individual and His Society. Ralph Linton was a col- 
laborator in Kardiner's seminar at Columbia University at the time 
this and subsequent books developed (e.g., Kardiner et al 1945, Du 
Bois 1944, West 1945) . He contributed ethnological reports based 
on his own earlier field work in the Marquesas in Polynesia, as well 
as in Madagascar. The high point of this undertaking was the study 
of the village of Atimelang on Alor, an island in eastern Indonesia, 
by Cora DuBois (1944). This was the first anthropological field 
work explicitly designed to employ an array of personality assess- 
ment techniques of psychologists in a non-European culture. These 
included the Rorschach, the Porteus Maze, word associations, chil- 
dren's drawings, autobiographies, and systematic observation of be- 
havior sequences. Most of these techniques had been used singly by 
earlier investigators, but their coordinated use was a distinct mile- 
stone. The methodological groundwork for the Kardiner-Linton 
collaboration was laid in Kardiner's earlier work, including seminars 
participated in by Ruth Benedict and Ruth Bunzel. Since prelimi- 
nary exploration of the method was worked out with Malinowski's 
Trobriand material, and the first full-scale analysis utilized Linton's 
Marquesan data, this collaboration can legitimately be claimed as 
Oceanic in origin. However, its impact was sufficiently great that 
it could not in any event be ignored in any review of the field. 

Although clearly psychoanalytic in his orientation, Kardiner rec- 
ognized cultural reality and cultural imperatives. Briefly, his analy- 
sis began logically with primary institutions, the cultural systems 
devoted to meeting essential needs. Adaptation and socialization in 
accordance with the dictates of the primary institutions requires the 
control of natural impulses. This control leads to frustration, and 
then to reactions to frustration, especially the formation of aggres- 
sive tendencies. The anxieties so created give rise to secondary insti- 
tutions, which are projections of anxiety in a variety of forms. The 
working out of anxiety is examined primarily at the level of the ego 
and of the superego in people, and through the analysis of projective 
systems in culture. 

This thumbnail summary obviously does not do justice to Kardi- 
ner's conceptual scheme of analysis, but it is sufficient to make clear 
the difference in emphasis in his approach from that adopted by 


Mead, or by Beaglehole. Both Mead and Beaglehole treat personality 
development in the broader framework of the learning of culture 
and its appropriate behaviors. Mead adds to this constitutional tem- 
perament and the effect of biological changes in maturation. Kardi- 
ner, in contrast, accounts for the same phenomena primarily in 
terms of psychological response to emotionally important experi- 
ences. In Kardiner's scheme the observable congruence in adult per- 
sonality necessarily requires the assumption that each individual 
who shares a culturally determined socialization experience will 
respond to it in substantially the same fashion as his fellows. Similar 
anxieties in a large number of people will then give rise to projective 
systems which serve to comfort them all. In the final section of this 
chapter we will return to an examination of this extremely crucial 

Without raising questions for the present regarding the useful- 
ness of either mode of analysis, the difference between Kardiner's 
approach on the one hand, and the emphasis partially shared by 
Mead and Beaglehole on the other, can perhaps be exemplified by 
parallel examples. Each deals with a culture in which older children 
have extensive responsibility for the care of their younger siblings 
during the day. The cultural behavior, and the reason for its exist- 
ence, are highly comparable in both instances, but the significance 
seen in it differs sharply. 

First, Kardiner's discussion of Alor {Kzr diner et al. 1945, p. 155) : 

In late childhood . . . both sexes are prematurely inducted into the role of taking 
care of their younger siblings. The performance of this role is undoubtedly subject 
to much variation. In general, however, a child who is robbed of the care essential 
for growth and development will not bestow such care upon a younger claimant 
without resentment. The result is that the older child, who is now the mother 
surrogate, is no more dependable than the mother herself. So the situation for the 
younger child is not greatly ameliorated by this institution. On the other hand, 
the older sibling is likely to be given attributes which were prevented expression 
toward the mother by the strong ambivalence to her. This attitude is furthermore 
facilitated by both older and younger sibling having a common claim. This is a 
factor which in some would tend to ameliorate the situations of sibling rivalry 
and render the hatred toward the parent still greater. In others it might terminate 
in intensified sibling rivalry and hatred. 

Contrast this with the view of James Ritchie, a student of Beagle- 
hole, of essentially the same behavior in Rakau (Ritchie 1956:47) : 

The Maori child is typing himself against an older sibling's concept of the 
adult world. His perceptions of adult behavior and adult roles are being strained 
through the perceptions of his older sib. The latter will only be approximately 


varying in their degree of conformity according to the age, sex, intelUgence and 
experience variables of the older child. In this transmission of percepts from a 
child's view of the world, the value structure is thrown into sharp relief. The 
limited comprehension of the older child requires that the values he sees around 
him be used in modifying the behaviour of younger children; he cannot there- 
fore make do with a tentative approximation but must resolve his percepts into 
a formal structure from which he is able to direct and instruct younger children. 
Originality departs. The value-structure sets hard, prematurely, and the child 
enters onto a plateau in value-learning. The organized model with which he has 
been presented will do for all situations right up to the time he assumes direct 
adult behaviour and even then a rigid conformity based on the simplicity and 
absolutism of the middle years will be a ready source of certainty in conflicting 
or incipiently dangerous social situations. 

Although Beaglehole's students, and indeed Beaglehole himself, are 
not always consistent in viewing personality as learned rather than 
as shaped by emotional response, they do represent a minority who 
are carrying forward and developing this approach. As the citation 
from Ritchie indicates, they are especially concerned with when, 
and from whom, a person learns his culture, his attitudes, and his 
ways of behaving. 

However, it is the work of Kardiner and his associates which one 
finds most commonly cited as methodological models for subsequent 
monographs in culture and personality. This is somewhat paradoxi- 
cal, because practically no one has been able to make effective use 
of his central concept of primary and secondary institutions. Ref- 
erence is made instead to some techniques he has employed — espe- 
cially the use of projective tests and life histories. This is perhaps the 
key to the paradox. Kardiner's books (1939, 1945) and DuBois' 
The People of Alor (1944) were available at the time, shortly after 
the war, when clinical psychologists in large numbers became inter- 
ested in cultural differences. Undoubtedly Kardiner's work spurred 
this trend, but its real impetus derived from the participation of 
psychologists in wartime intelligence analysis and psychological 
warfare. When psychologists then began to collaborate with and 
train anthropologists they found Kardiner's tools were the ones 
with which they were themselves familiar. Kardiner was not the 
first to use any of these tools, but he brought them together in a 
persuasive and effective manner. 

The psychologists were also comfortable in accepting Kardiner's 
assumption of the primacy in personality development of the in- 
dividual's intrapsychic integration of emotional experience. Yet 
while citing Kardiner to legitimize their focus on emotional deter- 


minants of behavior, the psychologists and their anthropological col- 
leagues disregarded the one solid tie to culture in Kardiner's scheme 
— his concept of primary and secondary institutions. The latter 
concept may or may not be useful. But the net effect has been an 
uncritical acceptance of both the theory and the tools of clinical 
psychology in culture and personality studies, an acceptance more 
wholehearted even than Kardiner's (cf. Hsu 1952, 1955). 

A final landmark can be identified with the late 1930's, John 
W. M. Whiting's Becoming a Kwoma (1941). This arose from a 
ferment of interest at Yale in the anthropological implications of 
the theories of learning and behavior developed by Clark Hull and 
his students {cf. Miller and Dollard 1941) , largely on the basis of 
learning experiments with rats. Whiting wrote a standard ethnog- 
raphy of the Kwoma, a mountain tribe in the Sepik River area of 
New Guinea, with considerable attention devoted to personality 
development. He then reanalysed his material in terms of drive, cue, 
response, and reward as an exercise in the application of Hull's the- 
ory of learning to a set of concrete ethnographic data. A brief exam- 
ple will suffice to illustrate the mode of analysis: 

In adolescence a boy learns to carry on secret love affairs with adolescent 
girls. The drives are sex, sex appetite, and anxiety (sex impells him to seek girls, 
sex appetite leads him to choose a girl culturally defined as attractive, and anxiety 
impells him to do so secretly) ; the response is the complex of behavior which 
leads to and includes sexual intercourse in the bush; the cues are the sight of an 
attractive girl, verbal permission from her, the environmental scene which has 
both public and secluded spots, etc.; the reward is sexual orgasm, satisfaction of 
sex appetite, and anxiety reduction, (pp. 176-177) 

Whiting's application of Hull's concepts to the Kwoma was so 
literal that it was almost a tour de force. The exercise has therefore 
not been repeated by others. But it was an instructive undertaking. 
It undoubtedly contributed to the explicit and scrupulous approach 
to theory which has since been characteristic of Whiting and his 
students. It also served to refine and make more effective the use 
of Hull's theory in culture and personality studies. 

During the early 1 940's much of Oceania became a theater of war. 
Field work necessarily ceased, and most anthropologists were in any 
event otherwise engaged. Monographs based on earlier field work 
were published during this period, but there was a break in the 
continuity of research effort. After the close of World War II, 
several people who had worked in Oceania earlier returned to the 
field; their work has been discussed above. But a new and more 


numerous generation of anthropologists also came into the area, 
among them quite a few interested in culture and personality. Their 
work differed in two important respects from that of their prede- 
cessors. One was a shift in locale. The majority of new field work 
was undertaken in Micronesia. Not only was this an area formerly 
almost entirely closed to anthropologists, it was also comprised pri- 
marily of the small insular communities characterized in the intro- 
duction to this chapter as ideal for some types of research. Many 
correspondingly small island societies in Polynesia had of course been 
studied in the past, but their cultural transfiguration through for- 
eign contact was generally much greater than in Micronesia. Fur- 
thermore, a good deal of money became available for field work in 
Micronesia, and this had the not surprising effect of tipping the 
scales in favor of doing research in this area. 

The other difference is more subtle, and hopefully will prove tran- 
sitory. This is a sharp reduction in the amount of methodological 
pioneering displayed by students of culture and personality in post- 
war Oceania. The account thus far has been highlighted by a series 
of "firsts," of often rather daring developments of new methods or 
new theories which have had a widespread impact on the field. Any 
field of study tends to crystallize as it matures, but culture and per- 
sonality theory has certainly not yet fully stabilized. New ap- 
proaches — the use of projective tests or photographic analysis, for 
example — have since the war had their primary development else- 
where and then been applied later with variations in Oceania. The 
remainder of this account will therefore be more brief and selective 
than that which has preceded, confined essentially to major mono- 
graphic contributions. Virtually all of these, excepting those already 
discussed which stem from the continuing activity of persons al- 
ready in the field in the 193 o's, are based on work in Micronesia. 

In particular, the Coordinated Investigation of Micronesian An- 
thropology (CIMA) , sponsored by the Pacific Science Board of the 
National Research Council, put a large number of anthropologists 
and related scientists into the field in 1947. Among these were four 
persons primarily interested in studying culture and personality: 
Joseph and Murray on Saipan, Spiro on Ifaluk, and the present 
writer on Truk. In addition Lessa, on Ulithi, had a strong secondary 
interest in the field. 

Alice Joseph and Veronica Murray ( 195 1 ) undertook to see how 
much useful information could be derived from a relatively short 
study of Chamorro and Carolinian children (and a few adults) on 


Saipan. Although both are physicians, Joseph in particular was al- 
ready well known to anthropologists for her work with the Hopi. 
In this study they placed primary reliance on projective and per- 
formance tests administered to one hundred children of each of the 
two ethnic groups. The Bender-Gestalt test was interpreted by its 
author, Lauretta Bender. The Rorschach, Arthur Point Perform- 
ance Scale II, and the Porteus Maze were treated exclusively by the 
authors, using conventional scoring and interpretive procedures. 
This study did not, therefore, make any new contributions to meth- 
odology. Nor can it be said to have validated a field procedure for 
economical personality delineation. Numerous subsequent studies 
in which ethnographic and projective interpretations have been 
compared for cross-validation show the clear danger of accepting 
projective test results at face value. In fact, the authors' own find- 
ings tend to confirm this danger. They conclude their discussion 
of the Rorschach results with the prediction that "either large scale 
antisocial behavior with unconscious self-destructive aims or death- 
like apathy might be expected from the younger generation." 
(p. 202) Bender found that the normal Saipanese Gestalt patterns 
corresponded to those found in confusional states elsewhere, and 
speculated whether "environmental influences can, in a people with 
strong primitive tendencies, produce a state of intellectual perplex- 
ity and disorientation which will manifest itself in a disturbance of 
Gestalt function similar to that produced by toxic influences." 
(p. 142) As of this writing the children so delineated now range in 
age from 18 to 30 years and thus far show no external evidence of 
crippling psychopathology. Actually, it is quite plausible to con- 
clude that differences in perceptual orientation and in style of cog- 
nitive thinking were responsible for almost the entirety of the 
response patterns the authors found so bizarre. This is not the ap- 
propriate place in which to examine the manner in which these 
differences can produce such distortion in the particular tests used. 
But one would certainly feel more comfortable had the authors 
addressed themselves to this possibility rather than accepting at face 
value conclusions based on interpretive criteria developed with 
European and American subjects. 

Melford Spiro's study of Ifaluk was undertaken in conjunction 
with the late Edwin Burrows. A number of projective and attitudi- 
nal measures were used, coupled with a full ethnography and psy- 
chological interpretations of individual and group behavior. 
Unfortunately, although Spiro has published a number of important 


theoretical papers based on this work (e.g., 1951, 1959, i960), only 
the ethnographic account has been published in full (Burrows and 
Spiro 1957) . It is therefore not possible to review the culture and 
personality study here. 

It should be noted that both Spiro (1959) and Joseph and Mur- 
ray in their book contributed substantially to the accounts available 
in the literature of psychotic personalities on non-European cul- 
tures. Spiro presented three detailed case studies, and Joseph and 
Murray ten short summaries, plus brief coverage of disorders of 
other kinds. 

The Truk study undertaken by myself and Seymour Sarason, a 
clinical psychologist (Gladwin and Sarason 1953), was also in- 
tended to develop a relatively quick method of personality assess- 
ment, aided in this case by the presence of other anthropologists on 
the team who covered areas not directly relevant to personality de- 
velopment. The method was an evolution of that used by Du Bois 
( 1944) on Alor. Rorschach and Thematic Apperception Tests were 
used in conjunction with life histories and some dreams of 23 in- 
dividuals selected to include both "average" and deviant persons. 
These data were combined with a standard ethnography, "Blind" 
interpretations of the Rorschachs and TAT's were undertaken by 
Sarason, using a clinical mode of interpretation rather than placing 
reliance, as is more customary in such studies, on scoring categories 
and frequencies. It was felt that a clinical interpretation, while ad- 
mittedly more subjective, permitted fuller exploitation of the ma- 
terial produced by the subjects. This procedure also made possible 
explicit examination of the ways in which culturally determined 
perceptual modes affected the response pattern in all subjects, a fac- 
tor which is obscured in interpretations based upon the scoring of 
responses. The interpretations in this study appeared to have con- 
siderable face validity. The methodology used here was perhaps more 
rigorous and self-conscious than that usually found in culture and 
personality studies, at least those which attempt to collate a variety 
of sorts of data. But essentially very little in this monograph is really 
new methodologically or theoretically. More recently, I have re- 
analysed some of my data in an attempt to define the cognitive 
structure of Trukese thinking (Gladwin i960). 

William Lessa undertook on Ulithi an even more abbreviated 
method than those described above (Lessa and Spiegelman 1954). 
He administered the Thematic Apperception Test to 99 persons 
well distributed by age and sex, and scored the resulting stories in 


accordance with procedures developed by William E. Henry. His 
psychologist collaborator, Marvin Spiegelman, then interpreted the 
results solely on the basis of the comparative frequencies of different 
responses in the various age and sex groups. This was the first time 
the TAT had been used in this way, and Lessa found a quite satisfy- 
ing congruence between Spiegelman's conclusions and those based 
on the ethnographic data. This congruence held throughout a wide 
range of behaviors, including general motivational structure, han- 
dling of aggression, attitudes toward sex, food anxieties, etc. 
This does not necessarily mean that the TAT can be assumed to yield 
valid personality measures when used as a basis for quantitative in- 
terpretation in a culture other than Ulithi. However, an accumula- 
tion of similar evidence might encourage those anthropologists who 
are still interested in using projective tests to shift their emphasis 
away from the Rorschach. Interpretation of the Rorschach in any 
setting necessarily requires more inference than the TAT because 
the Rorschach presents a less structured stimulus. Rorschach inter- 
pretation rests on a larger series of assumptions about unconscious 
psychological processes derived from clinical experience In our own 
culture than does the TAT. Its cross-cultural application is there- 
fore inherently more hazardous — although by no means necessarily 
invalid If the instrument is used with due caution. 

With these few Micronesian studies by newcomers to the field, 
our review can be considered completed. Increasingly, of course, 
persons primarily Interested in other ethnological specialties none- 
theless Include psychological constructs In their observations and 
hypotheses, but this Is true of anthropology as a whole and need not 
be detailed for Oceania. We may therefore turn to an assessment of 
the work here reviewed within the broader perspective of the field 
of culture and personality as a whole. 


Oceania, as noted at the outset of this chapter, has held out an al- 
most unparalleled opportunity and challenge for students of culture 
and personality. We have seen the challenge well met. Much pio- 
neering field work has taken advantage of the unusual research 
settings afforded by island populations, and the data so derived have 
Inspired a number of bold but cogent new theoretical modes of In- 
terpretation. It would probably be justifiable to claim some of these 
new concepts as genuine anthropological contributions, in the sense 
that they could only have arisen from the necessity for explaining 


differences between cultures. Margaret Mead's exploration of the 
interrelationships between personality and social structure and 
maturational levels might be an example. The examination by Bea- 
glehole, Ritchie, et al., of the mechanisms of transmission of the 
psychological aspects of culture from one generation to the next is 
another. We might also cite the development by Porteus of the Maze 
Test for cross-cultural use, and Bateson's concept of schizmogenesis. 
Each of these approaches I would judge has the potential of making 
really new contributions to personality theory. 

However, few if any of them are to be found in the main stream 
of culture and personality theory as it is taught and used today. The 
field has come to place almost exclusive reliance upon theory and 
psychodynamic concepts derived from clinical psychiatry and psy- 
chology. The clinician undertakes to cope with and modify the emo- 
tional disturbances he finds in his patients, and he can understand 
the source and nature of these disturbances most readily by recourse 
to explanatory concepts derived from psychoanalysis. Years of 
research and clinical experience have modified, enriched, and elabo- 
rated this theoretical system since it was first set forth by Freud. The 
psychiatrist or psychologist finds here a handy and effective set of 
tools. Although he is urged to tinker with the system and perhaps 
improve a little upon it, he sees no reason to question its basic 

It is not surprising that anthropologists, seeing psychology un- 
animous in its support of a coherent and rather glittering body of 
theory, should accept from psychology the promise that this is the 
way to describe and account for personality and its development. 
However, the brief historical review just completed shows clearly 
that the relative unanimity on this score among anthropologists is 
of rather recent origin. I have suggested that the crystallizing event 
was the publication of Abram Kardiner's first book based on his 
collaboration with Ralph Linton. Whether or not others would 
agree upon this landmark, I believe few would disagree that psycho- 
analytic theory (as it is interpreted by clinical psychologists) now 
dominates the thinking of anthropologists in culture and personal- 
ity {cf. Kluckhohn 1944:590). It would therefore seem most ap- 
propriate in this assessment to take a second look at some of the 
earlier divergent views which arose from field work in Oceania, 
to see whether we have perhaps not been overly hasty in brushing 
them aside in our rush to leap upon the Freudian bandwagon. 

The most thoroughly documented and elaborated position is that 


of Margaret Mead, in many respects seconded as we have seen by 
Ernest Beaglehole. Mead does not, of course, reject the concepts of 
current personahty theory. In fact, she draws heavily upon them 
and has ever since her earhest work, done at a time when most an- 
thropologists were scarcely aware of the existence of a man named 
Sigmund Freud. However, analytic theory has two aspects. First, 
it is a conceptual scheme for the description of the emotional struc- 
ture of personality, of the forces within the individual which shape 
his behavior. But beyond this it also embraces a developmental 
scheme which undertakes to account for the formation, primarily 
in early childhood experience and on an emotional basis, of the psy- 
chological forces it has described. In psychiatry these two compon- 
ents of theory are thoroughly intertwined, to such a degree that the 
formal diagnostic criteria for most disorders require both behavioral 
and etiological determinations. However, Mead has in effect drawn 
upon the descriptive concepts while at the same time placing very 
minor reliance upon the developmental theory. 

Although Mead's difference is quite fundamental, and consistent 
throughout her work, it has probably received so little explicit at- 
tention largely because she has been content to be quietly selective 
rather than to attack the Freudian developmental premises. She has 
elaborated her own position while coexisting peacefully with the 
clinicians, being satisfied to show them the importance of cultural 
differences without attempting to force them to alter their basic 
theory. In her early work in Samoa and the New Guinea area she 
demonstrated the great differences in adult personality which result 
from growing up and living in different kinds of social environ- 
ments. Although she noted the stresses and strains of socialization in 
various types of life experience, she tended not to evaluate the im- 
portance of such crises or conflicts in terms of the significance at- 
tributed to them by a predetermined theoretical system. Rather she 
examined their consistency and congruence with other experiences 
which preceded, followed, and surrounded the situation under con- 
sideration. As has already been pointed out, this approach views 
personality as a system of thinking, feeling, and behaving which is 
learned through continuing experience. In her view, people learn to 
conform to the norms not merely to avoid punishment or gain re- 
wards, but also because in this way life becomes more predictable 
and meaningful. 

The learning process is central to Mead's scheme, and by learning 
she means not merely the factors which may stimulate and affect 


the permanence of learning, but also the context and content of 
learning. Learning for Mead is therefore a more inclusive and de- 
scriptive concept than it is in the learning theory derived from Hull, 
but by the same token less subject to experimental manipulation 
and verification. In her scheme, the learning which culminates in 
adult personality is the end product of an infinity of small experi- 
ences shared with other people. The process is consequently much 
more difficult to capture and define than one which is postulated 
as a response to a more limited array of emotionally charged critical 
situations or relationships. In her early work Mead was content to 
describe the social environment and the manner in which it was 
interpreted by the individual and by the persons about him, leaving 
unexamined the detailed process whereby this cultural transmission 
took place. In Bali she tried to actually document the process, largely 
through photography, and in one study already referred to (Mead 
and Macgregor 1951), she undertook to spell out all the nuances 
of learning of one component of behavior. 

I have risked redundancy in restating here the development of 
Margaret Mead's theoretical position in order to underscore the con- 
sistency of its development, and its completeness. She has not only 
a point of view, but also a research method consistent with her point 
of view. Furthermore, this is a point of view which is essentially an- 
thropological. Personality, to Mead, is part of the cultural heritage 
to be passed on from one generation to the next. It is learned by each 
generation in much the same way as is canoe building, speaking, 
or social etiquette. Because it is learned, and because it is learned 
through living the culture, it necessarily develops, with variations, 
in essentially similar form from one person to the next. No two in- 
dividuals in a given society are identical in the way they build a 
canoe or in the way they feel toward their mothers, but within each 
society everyone does these things, and many others, in a fashion 
sufficiently uniform and distinctive to be characteristic of the cul- 
ture they share. If people did not learn to behave with this essential 
uniformity, anthropologists would not be warranted in speaking 
of culture and cultural differences. Mead in effect sees no reason why 
anthropologists should then not consider personality simply as an- 
other component of culture, to be studied as far as possible in the 
same way. 

Let us now examine briefly the contrasting but currently more 
popular view, that represented by the personality theory of the clini- 
cal students of human behavior. This approach, as befits one tailored 


in the first instance to the needs of individual patients, stresses the 
integration of the personahty within the individual. This integra- 
tion is developed through adaptive response to experience as it is 
emotionally perceived and interpreted by the individual. People 
learn to behave and feel in certain ways because this will defend 
them from anxiety or other distressing psychological experiences, 
or will bring them love and reward. With this view is associated a 
conviction that the experiences of early childhood are more crucial 
and lasting in their impact than later ones. However, it should be 
noted that this emphasis on early experience is in no way theoreti- 
cally required by the primary focus on emotional integration {cf. 
Hsu 1952). 

Studies in culture and personality which use this scheme account 
for the similarities of adult personality found in a given society by 
the culturally determined similarities in (early) socialization ex- 
perience. As stated, this does not differ from Mead's scheme. The 
difference lies in the fact that in the psychoanalytic framework the 
intervening operative variable is emotional integration within each 
developing individual. Thus, for example, in a society in which 
adults characteristically reveal strong anxieties about food in excess 
of the actual danger of going hungry, and where the nursing of ba- 
bies is inconsistent or otherwise frustrating, analytic theory might 
lead to explanations of adult anxieties in terms of unfulfilled oral 
needs (despite the lack of thumb sucking observed in several such 
societies) . In contrast Mead would probably also say that if young 
children see strong capable adults worried about food, they are most 
likely to learn to worry also. She would look at the social structure, 
perhaps finding that the mutual responsibilities of kin groups are 
so arranged that food which is objectively obtainable at all times is 
in fact frequently hard to get. And she would look at the biological 
rhythms and nutritional needs of the people to inquire whether 
these created special conditions or problems. Similarly, the com- 
monly felt fear of heights in our culture can be attributed either, on 
the one hand, to a symbolic fear of loss of support by a loved person, 
or, on the other, to the fact that mothers in our society usually show 
a panic reaction when they find their children climbing trees, build- 
ings, etc., and thus teach the child to be frightened of falling. There 
may even be inherited differences in different population groups 
with respect to perception and balance, as these affect reactions to 

With this outline of the differences between the available meth- 


odological strategies before us, how can we assess their respective 
vahdity and utihty? No definitive evaluation is possible, and what 
follows must necessarily represent only my own opinion. 

The research objective of culture and personality studies employ- 
ing conventional analytic personality theory is essentially to test 
the validity of the concepts subsumed under that theory in a variety 
of cultural contests. A plausible explanation of some aspect of per- 
sonality development in another culture serves to buttress the va- 
lidity of the particular explanatory formulation derived from our 
own culture. If our explanation does not "fit" the other culture, the 
original concept must be discarded or reinterpreted. Anthropology 
becomes the handmaiden of psychology, testing a secondhand the- 
ory without any real opportunity to lead the way to new and differ- 
ent understandings of personality development.^ 

Irrespective of the productivity of using analytic theory in cul- 
ture and personality research, the more serious question of validity 
must be examined. In the first place, it should be noted that the 
interpretative substance of a monograph written in this vein is not 
the exposition of an observed ongoing process such as anthropolo- 
gists usually favor. Rather it is a series of post hoc explanations of 
developmental processes, working back from adult personality. The 
plausibility of such interpretations can be attributed equally to their 
inherent validity or to the ingenuity of the explainer. Post hoc inter- 
pretations are of course appropriate to the theory used because 
psychoanalytic case studies, upon which the theory has largely been 
built, are usually of adults and therefore retrospective. However, 
insofar as this theory is clinically validated, its validation rests upon 
a successful therapeutic outcome. Such an outcome is likely to de- 
pend as much or more upon the skill of the therapist than upon the 
accuracy of the theory. A developmental history which will serve 
well in therapy will not necessarily serve science. 

If we look at these same clinical tools in the hands of those psychi- 
atrists who have elected to work with children, the picture becomes 
quite different. It must be remembered that, as used in culture and 
personality studies, analytic theory posits that similar early life ex- 
periences are integrated similarly by a large number of people to 
produce a distinctive adult personality common to all of them. Yet 
no responsible child psychiatrist, even when faced with a young 

* For what interest it may have, it might be mentioned that some years ago I discussed 
precisely this same research approach and arrived at an opposite and much more enthusiastic 
conclusion (Gladwin and Sarason 1953:21-22). 




patient who has experienced several clearly traumatic years, will 
use his clinical concepts to predict this child's adult personality with 
nearly the precision which is taken for granted in accounting for 
culturally determined basic personality structure. The child psy- 
chiatrist may feel safe in saying the child will probably always be 
maladjusted. But he would consider it foolhardy in any one case, to 
say nothing of hundreds, to state in just what form the child's 
anxieties will become crystallized, how his defense mechanisms and 
projective systems will be structured, what sorts of behaviors this 
will lead him to adopt in a variety of adult situations, and so on. In 
other words, the same body of clinically derived theory which per- 
mits psychiatrists to make post hoc explanations for therapeutic 
purposes becomes unthinkable, even in a clinical context, as a basis 
for the very sort of prediction of outcomes of childhood experience 
which are essential to their valid use in culture and personality 

Therefore the comparison between, on the one hand, Margaret 
Mead's view of personality as simply an aspect of culture and biol- 
ogy, and on the other, the more analytically oriented view of most 
workers in the field today, leads, in my opinion, to a discouraging 
view of the latter. Stated in extremes, we have surrendered our an- 
thropological birthright to the clinicians, and received in return a 
methodology which is both limited in productivity and suspect in 
validity. Obviously, the situation is not that bleak. If nothing else, 
the work done thus far has provided a thorough exploration of one 
approach, and has unquestionably served to enrich the theory it has 
borrowed. Increasingly, however, not only anthropologists, but also 
psychologists and sociologists, are wondering where all this work 

If the answer is to be hopeful, it is my conviction that anthropol- 
ogists must be prepared to make a commitment to their own theory 
of culture as full as their present commitment to the psychologists' 
theory of personality. As Hsu (1955) has observed, the predictable 
similarity in behavior between members of any single society has 
been noted by travelers ever since Herodotus. This striking phe- 
nomenon obviously cannot be accounted for solely by the psycho- 
dynamics of development within each individual in that society. It 
is a cultural phenomenon, and anthropologists must view it as such. 
It is of the same order as similarities in house types or agricultural 
methods. This, of course, does not mean that psychological theory 
should be discarded. One cannot speak of houses without attention 


to architectural principles, or of agriculture without considering 
the chemistry of soils and nutrition. What is needed is real collabora- 
tion, not one-sided borrowing, in the relationship between anthro- 
pology and psychology. 

One other aspect or component of personality remains to be con- 
sidered: cognitive process, or the style of thinking and problem solv- 
ing which characterizes a culture. Cognitive development has been 
almost entirely overshadowed in culture and personality research 
by the emphasis on emotional development. This is doubtless in part 
because of a heavy reliance on analytic theory in which cognition 
plays a very small role. But it is also because psychology does not 
itself have an agreed upon body of cognitive theory. Anthropology 
has edged into this field of inquiry largely through linguistics. Ex- 
plicit attention to cultural differences in logical process as such is 
rare in anthropology. Surprisingly, almost all of this work has been 
based on data from Oceania, and has been cited above (Bateson 
1942, 1958; Beaglehole 1957; Gladwin i960; Lee 1950; Porteus 
1931, 1937; see also Margaret Mead's [1932] analysis of Manus 
animism) . 

In 1936, in the first edition of Naven, Bateson emphasized the 
distinctness and importance of cognitive processes. He character- 
ized the usual grist for the culture and personality mill as ethos, "the 
expression of a culturally standardized system of the organization 
of the instincts and emotions of . . . individuals" (1958:118). Com- 
plementary to ethos is eidos, "a standardization [and expression in 
cultural behavior] of the cognitive aspects of the personality of 
individuals" (p. 220) . Eidos embraces such matters as the nature 
of memory, the perception and structuring of external reality, the 
possibility of a positive valuation of intellectuality (e.g, expert 
knowledge of genealogy or folklore) , and preferred strategies in 
problem solving. Subsequently, he carried eidos one step farther, 
evolving the concept of deutero-learning, or learning how to learn, 
referring to the context or intellectual tools of learning (Bateson 
1942). In neither instance did Bateson carry through with a full 
review of his ethnographic material to demonstrate the potentiali- 
ties of an analysis in these terms. His concepts, however, are impor- 
tant in that they point to culturally determined differences in the 
basic intellectual tools available to persons reared in different so- 
cieties. When one remembers that Naven first appeared over 25 
years ago, it is hard to understand why so few have been moved to 
pick up this line of inquiry. 


The significance of Bateson's insight, and also the fact that it 
could come only to an anthropologist trying to account for ob- 
served cultural differences, is apparent if one looks at the efforts of 
psychologists to grapple with the nature and development of in- 
tellect. The often brilliant studies of such psychologists as Bartlett, 
Bruner, Guilford, Hebb, Piaget, and others, have one characteristic 
in common: they make the assumption that real intelligence con- 
sists in the ability to integrate information in symbolic and rela- 
tional terms, and thus subsume large amounts of data through 
abstract generalizing principles. This assumption is entirely reason- 
able for persons working within our culture. Virtually all our ma- 
jor intellectual achievements are predicated upon just this mode of 
abstract thinking. 

However, as anthropologists we must raise the question whether 
it is not culturally parochial to view abstract thinking as the only, 
or even the best, form of intelligence. Here we may refer to Beagle- 
hole's conclusion, cited earlier, that the Aitutaki do not think ab- 
stractly, and in fact do not value problem -solving ability in the 
terms we know it, that is, in terms of conscious rational processes. 
Yet they and their forebears have developed a complex and adaptive 
technology. To mention another example, Sarason and I found the 
Trukese to have a highly concrete nonabstract style of thinking.* 
Yet the Trukese also not only have a very useful technology, but can 
be demonstrated in their highly evolved techniques of interisland 
navigation to be accomplishing entirely in their heads some truly 
extraordinary feats of data reduction and problem solving (Glad- 
win i960) . The Trukese navigator is clearly not equipped to em- 
brace the logical systems analysed and studied, for example, by 
Piaget, but it is hard to say that he is not being distinctly intelligent. 

Piaget, of course, and the other psychologists mentioned are not 
studying the totality of intelligence, but only intelligence-in-our- 
culture.^ Or, to be more correct, they are studying intelligence-as- 
valued-by-middle-class-intellectuals-in-our-culture. The latter 
phrasing points up the problem more sharply. Even if he had avail- 
able a detailed analysis of Trukese thinking it is doubtful that Piaget 
would or should change his research approach. It is in the evaluation 

* The distinction between concrete and abstract thinking refers, of course, to differences in 
emphasis in basic problem-solving strategy. It is doubtful that any thought process could be 
totally concrete or totally abstract, 

^ For views of Piaget, Margaret Mead, and others on this issue, see World Health Organization 


of the full range, rather than the upper reaches, of intelligence that 
a cross-cultural perspective can have the greatest impact. 

Psychologists essentially concern themselves with only one cri- 
terion of fully developed intelligence — abstract symbolic manipu- 
lation of information. They therefore tend to measure — and indeed 
define — intelligence in terms of tests which at each higher age level 
require more ability in abstraction. It is disturbingly suggested by 
work in our own culture, and often obvious when the tests are tried 
out in other cultures, that some people are ill equipped to cope with 
our intelligence tests even though they can meet the mental prob- 
lems posed by their culture and environment with assurance and 
success. Furthermore, identifiable groups of people (e.g., lower-class 
Italians in the U.S.) have characteristic sorts of difficulties with tests 
(cf. Masland, Sarason, and Gladwin 1958, chap. 14) . Psychologists 
have been troubled by this situation, and have attempted to develop 
a variety of culturally fair tests. Best known are the Davis-Eells 
Games, and the Cattell Culture-Free Test. However, what they have 
generally done is to make the content revolve about familiar situa- 
tions and reduce or eliminate the explicit verbal skills required, 
while leaving essentially intact the kind of reasoning ability required 
for effective performance. 

Meanwhile, anthropologists have done little to help them other 
than to insist piously that all groups and classes of men, regardless 
of cultural origin, must have equal intellectual potentialities. 
Anthropologists have contributed very little toward giving psy- 
chologists an understanding of the meaning of intelligence-in-our- 
culture as against intelligence-in-another-culture. With respect to 
emotional factors, personality and culture studies have assuredly 
given psychologists a valuable perspective. As a consequence, psy- 
chologists feel comfortable in looking for rather fundamental dif- 
ferences in personality development in the various subcultures of 
our society. Quite aside from the theoretical importance of cogni- 
tive theory, it is high time anthropology lent a similar helping hand 
to psychologists in the study of thinking and intelligence. Psychol- 
ogy is in the troubled position of lacking the theoretical and practi- 
cal tools to disprove the racial inferiority its own tests are constantly 
being cited to "prove." It appears that only a wide cross-cultural 
perspective can provide a foundation of knowledge upon which to 
develop such tools. 

Summing up, it seems fair to say that the challenge of unusual 
research opportunities offered by Oceania has indeed proved stimu- 


lating. As was noted, practically the entirety of explicit anthropol- 
ogical contributions to the study of cognitive process has stemmed 
from this area. Two major approaches to a genuinely anthropologi- 
cal — i.e, cultural — theory of personality development have been 
developed by Mead and by Beaglehole and their colleagues. Yet there 
seems at present to be a slackening in leadership in Oceania. Hope- 
fully this is illusory, or at least temporary. The challenging oppor- 
tunities remain. The relationship between personality and culture 
change, mentioned earlier in this chapter, is only one of several lines 
of inquiry which have scarcely been exploited at all, but which can 
fruitfully be pursued in the Pacific area. 


Bateson, Gregory 

1942 Social planning and the concept of "deutero-learning." Conference on 
Science, Philosophy, and Religion, Second Symposium. New York, 
Harper & Bros. 
1949 Bah: the value system of a steady state. In, Social structure: studies pre- 
sented to A. R. Radcliffe-Brown. Meyer Fortes, ed. 35-53. Oxford, 
Clarendon Press. 
1958 Naven. id ed. Stanford, Stanford University Press. 
Bateson, Gregory and Margaret Mead 

1942 Balinese character: a photographic analysis. Special Publications of the 
New York Academy of Sciences II. 
Beaglehole, Ernest 

1939 Some modern Hawaiians. University of Hawaii Research Publica- 
tions 19. 

1940 Psychic stress in a Tongan village. Proceedings of the Sixth Pacific 
Science Congress. Berkeley, University of California Press 4:43-52. 

1 94 1 Pangai village in Tonga. Memoirs of the Polynesian Society 18. 

1944 Character structure: its role in the analysis of interpersonal relations. 

Psychiatry 7:145—162. 
1957 Social change in the South Pacific: Rarotonga and Aitutaki. New York, 

Beaglehole, Ernest and Pearl 

1935 Hopi of the Second Mesa. Memoirs of the American Anthropological 

Association 44:1-67. 
1938 Ethnology of Pukapuka. Bulletins of the Bernice P. Bishop Museum 

1 94 1 Personality development in Pukapukan children. In Language, culture, 

and personality. Leslie Spier, A. Irving Hallowell, and Stanley S. 

Newman, eds., pp. 282—298. Menasha, Sapir Memorial Publication Fund. 
1946 Some modern Maoris. New Zealand Council for Educational Research. 

Wellington, "Whitcombe and Tombs. 


Beaglehole, Ernest and James E. Ritchie 

1958 The Rakau Maori studies. Journal of the Polynesian Society 67: 132-154. 
Benedict, Ruth 

1934 Patterns of culture. New York, Houghton Mifflin. 

1938 Continuities and discontinuities in cultural conditioning. Psychiatry 

Berndt, R. M. and Catherine H. 

195 I Sexual behavior in Western Arnhem Land. Viking Fund Publications 
in Anthropology No. 16. New York, Wenner-Gren Foundation for 
Anthropological Research, Inc. 

Burrows, Edwin G. and Melford E. Spiro 

1957 An atoll culture: ethnography of Ifaluk in the Central Carolines. 2d ed. 
Behavior Science Monographs. New Haven, Human Relations Area 

Du Bois, Cora 

1944 The people of Alor: a social psychological study of an East Indian 
island. Minneapolis, University of Minnesota Press. 

Earle, Margaret Jane 

1958 Rakau children: from six to thirteen years. Wellington, Victoria Uni- 
versity of Wellington Publications in Psychology 11. (Monographs on 
Maori Social Life and Personality 4.) 

Fortune, Reo F. 

1927 The mind in sleep. London, Kegan Paul. 

1932 Sorcerors of Dobu: the social anthropology of the Dobu islanders of the 
Western Pacific. New York, Dutton. 
Gladwin, Thomas 

i960 The need: better ways of teaching children to think. In Freeing Capac- 
ity to Learn: papers and reports from the Fourth ASCD Research 
Institute, pp. 23—39. Washington, National Education Association. 
Gladwin, Thomas and Seymour B. Sarason 

1953 Truk: man in paradise. Viking Fund Publications in Anthropology No. 
20. New York, Wenner-Gren Foundation for Anthropological Re- 
search, Inc. 

Hsu, Francis L. K. 

1952 Anthropology or psychiatry: a definition of objectives and their im- 
plications. Southwestern Journal of Anthropology 8:227-250. 
1955 An anthropologist's view of the future of personality studies. Psychi- 
atric Research Reports 2:155—168. 

Joseph, Alice and Veronica F. Murray 

195 I Chamorros and Carolinians of Saipan, personality studies. Cambridge, 
Harvard University Press. 

Kardiner, Abram 

1939 The individual and his society: the psychodynamics of primitive social 
organization. New York, Columbia University Press. 

Kardiner, Abram, et al. 

1945 The psychological frontiers of society. New York, Columbia Uni- 
versity Press. 


Kluckhohn, Clyde 

1944 The influence of psychiatry on anthropology in America during the 
past one hundred years. In One hundred years of American psychiatry, 
pp. 589—617. New York, Columbia University Press. 
La Barre, Weston 

1958 The influence of Freud on anthropology. American Imago 15:275-328. 
Lee, Dorothy M. 

1950 Lineal and nonlineal codifications of reality. Psychosomatic Medicine 
Lessa, William A. 

1956 Oedipus-type tales in Oceania. Journal of American Folklore 69:63—73. 
Lessa, William A. and Marvin Spiegelman 

1954 Ulithian personality as seen through ethnological materials and the- 
matic test analysis. University of California Publications in Culture 
and Society 2:243—301. 

Malinowski, Bronislaw 

1922 Argonants of the Western Pacific, an account of native enterprise and 

adventure in the archipelagoes of Melanesian New Guinea. London, 

1927 Sex and repression in savage society. London, Kegan Paul. 
1929 The sexual life of savages in North-Western Melanesia. New York, 

1935 Coral gardens and their magic, a study of the methods of tilling the soil 

and of agricultural rites in the Trobriand Islands. New York, American 

Book, 2 vols. 

Masland, Richard L., Seymour B. Sarason and Thomas Gladwin 

1958 Mental subnormality: biological, psychological, and cultural factors. 
New York, Basic Books. 

Mead, Margaret 

1932 An investigation of the thought of primitive children, with special 
reference to animism. Journal of the Royal Anthropological Institute 

1937 Cooperation and competition among primitive peoples. New York, 

1939 From the South Seas: studies of adolescence and sex in primitive so- 
cieties. New York, Morrow. (Comprises reprints of Coming of age 
in Samoa, 1928; Growing up in New Guinea, 1930; Sex and tempera- 
ment in three primitive societies, 1935.) 

1949 Male and female: a study of the sexes in a changing world. New York, 

1953 National character. In Anthropology today, A. L. Kroeber, ed., pp. 
642—667. Chicago, University of Chicago Press. 

1954a Research on primitive children. In Manual of child psychology. Leonard 
Carmichael, ed., 2d ed., pp. 735-80. New York, Wiley. 

1954b Cultural continuities and personality transformation. Journal of Social 
Issues, Supplement Series No. 8. Kurt Lewin Memorial Award Issue. 

1956 New lives for old, cultural transformation: Manus 1928-1953. New 
York, Morrow. 


1959 Reply to Kardiner. Science 130:1728, 1732. 
Mead, Margaret and Frances Cooke Macgregor 

195 1 Growth and culture: a photographic study of Balinese childhood. New 
York, Putman's. 
Mead, Margaret and Rhoda Metraux 

1953 The study of culture at a distance. Chicago, University of Chicago 
Mead, Margaret and Theodore Schwartz 

i960 The cult as a condensed social process. In Group Processes: Transactions 
of the Fifth Conference, Oct. 12, 13, 14 and 15, 1958, Princeton, N.J. 
Bertram Schaffner, ed.. New York, Josiah Macy, Jr. Foundation, 
Mead, Margaret and Martha Wolfenstein, eds. 

1955 Childhood in contemporary cultures. Chicago, University of Chicago 

Miller, Neal E. and John Dollard 

1 94 1 Social learning and imitation. New Haven, Yale University Press. 
Mulligan, D. G. 

1957 Maori adolescence in Rakau. Wellington, Victoria University College 
Publications in Psychology 9. (Monographs on Maori Social Life and 
Personality 2.) 
Myers, Charles J. and W. McDougall 

1903 Reports of the Cambridge Anthropological Expedition to Torres Straits 
II, Physiology and Psychology. Part 2. Cambridge, Cambridge Uni- 
versity Press. 

193 I The psychology of a primitive people: a study of the Australian Ab- 
origine. New York, Longmans, Green. 
1937 Primitive intelligence and environment. New York, Macmillan. 

Ritchie, James E. 

1956 Basic personality in Rakau. Wellington, Victoria University College 
PubUcations in Psychology 8. (Monographs on Maori Social Life and 
Personality i.) 

Ritchie, Jane 

1957 Childhood in Rakau: the first five years of life. Wellington, Victoria 
University College Publications in Psychology 10. (Monographs on 
Maori Social Life and Personality 3.) 

Rivers, W. H. R. 

1923 Conflict and dream. New York, Harcourt, Brace. 

Roheim, Geza 

1925 Australian totemism: a psycho-analytical study in anthropology. 

London, Allen and Unwin. 
1932 The psychoanalysis of primitive cultural types. International Journal 

of Psychoanalysis 13:1—224. 
1950 Psychoanalysis and anthropology. New York, International Universities 



Spiro, Melford E. 

195 1 Culture and personality: the natural history of a false dichotomy. Psy- 
chiatry 14:19—46. 
1959 Cultural heritage, personal tensions, and mental illness in a South Sea 
culture, bt Culture and mental health, Marvin K. Opler, ed., pp. 141- 
171. New York, Macmillan. 
1961 Social systems, personality, and functional analysis. In Studying person- 
ality cross-culturally, Bert Kaplan, ed. 
Taylor, C. R. H. 

195 1 A Pacific bibliography. Wellington, N. Z., The Polynesian Society. 
ViNACKE, W. Edgar, et al. 

1955 Bibliography of sources on personality and culture of the Pacific region. 
Honolulu, University of Hawaii (mimeographed). 
West, James 

1945 Plain ville, U.S.A. New York, Columbia University Press. 
Whiting, John W. M. 

194 1 Becoming a Kwoma: teaching and learning in a New Guinea tribe. New 
Haven, Yale University Press. 
Williams, John Smith 

i960 Maori achievement motivation. Wellington, Victoria University of 
Wellington Publications in Psychology 13. (Monographs on Maori So- 
cial Life and Personality 5.) 
World Health Organization 

1957 Proceedings of the Fourth Meeting, Study Group on Psychobiological 
Development of the Child, Sept. 20—26, 1956, Geneva. Mimeographed. 
(To be published by Tavistock Press.) 

Chapter 6 


Harvard University 

The method of analysis which yields studies in "culture and per- 
sonality" when applied to "primitive" peoples has its analogue 
among studies of large-scale societies in a varied assortment of in- 
vestigations on what is called national character. If, under this head- 
ing, we allow impressionistic, introspective, and loosely evaluative 
works to qualify, then for the United States alone — from De Toc- 
queville to Brogan and Gorer — the articles and books depicting the 
American character will be numbered in the hundreds (Commager 
1947). Were we to extend our coverage to the major nations of 
Europe and Asia, the number of relevant studies would be in the 
thousands. To review even the most important of these would strain 
the limits of our allotted space even while permitting only the driest 
catalogue of their contents. Yet if we were to insist on the more rig- 
orous standards of empirical social science, and were to consider 
only more systematic investigations based on representative samples 
and utilizing standard psychological tests, then not more than two 
or three studies in the relevant literature could qualify. There is a 
third alternative. By selecting a specific problem focus we may si- 
multaneously escape the boundlessness of a general review and the 
confining restrictions forced on us through the adoption of a rigor- 
ous methodological canon. A topic suitable to our purpose, one of 
interest and importance, is the relation of national character to the 
political systems found in modern national states, and more spe- 
cifically, to the establishment and maintenance of democracy. Be- 

* Revised and expanded version of a paper read at the Fourth World Congress of Sociology, 
Stresa-Milan, 1959. The aid of the Social Science Research Council is gratefully acknowledged, 
as well as the support of the Russian Research Center at Harvard. Professors S. N. Eisenstadt 
and Daniel J. Levinson were kind enough to offer numerous excellent suggestions. 





fore we examine this relationship, we must clarify the meaning of 
our concepts. 


Problems of Definition 

The confusion about the term national character is pervasive and 
enduring. Yet arguing about what a concept should mean can be 
utterly sterile. What is important is that we designate some empirical 
phenomenon which has concrete reference, which can be effectively 
distinguished from other phenomena, and which can conceivably 
be investigated by standard replicable, reliable, and valid methods. 
For purposes of this discussion I will adopt the definition of national 
character presented in the Handbook of Social-Psychology (Inkeles 
and Levinson 1954) which, I believe, is now widely accepted: "Na- 
tional character refers to relatively enduring personality charac- 
teristics and patterns that are modal among the adult members of 
a society." 

The other meanings given to national character, and related 
terms such as people's character, folk character, national (or "ra- 
cial" or popular) psychology, are almost as numerous as the roster 
of political essayists from Plato to Pareto and from Pareto to Potter. 
Some treat national character as simply "the sum total" of all the 
values, institutions, cultural traditions, ways of acting, and history 
of a people. However useful this idea may be for popular discourse, 
it is sadly lacking for purposes of scientific analysis, since the failure 
to differentiate the elements of the phenomenon makes an impos- 
sible task of measurement, obfuscates issues of cause and effect, and 
precludes systematic study of the relations between elements. With 
most other definitions we have no quarrel, so long as those using the 
different terms are appropriately aware that each has a special and 
restricted meaning, and that no one of these concepts exhaustively 
describes the phenomenon under investigation. The following main 
types of definition may be discerned (cf. Herz 1944, and Kline- 
berg 1944) : 

National Character as Institutional Pattern. In this approach, 
most common among political scientists, the national character is 
epitomized by the dominant, or typical and representative, institu- 
tions, particularly those concerned with politics and economics. 
The choice between dominant as against typical or representative 



institutions as the basis for characterizing a nation is a difl&cult one, 
and has led to much confusion in those studies in which the dis- 
tinction was not precisely made or rigorously adhered to. Outstand- 
ing examples of the genre are to be found among numerous studies 
of the American character, such as those by Andre Siegfried (1927) 
orD. W. Brogan (1933, 1944). 

National Character as Cultiire Theme. Broadly similar to the 
preceding approach, this genre gives prime emphasis not to political 
and economic institutions but to the family, friendship, the local 
community, and to values, attitudes, philosophy of life, religion and 
the like. Themes are often selected as cutting across or as infusing 
these and other social realms. Most common among anthropologists, 
this approach is also typical for many historians, political scientists, 
and essayists who speak in terms of spirit or folkgeist, world out- 
look, life-ways, and similar themes. Perhaps the best known of the 
more or less modern efforts of this type would be de Madariaga's 
Englishmen, Frenchmen, Spaniards (1929) , and the most impres- 
sive of the recent statements, Ruth Benedict's The Chrysanthemum' 
and the Sword ( 1946) . 

National Character as Action. In this approach stress is placed 
on behavior and its consequences, with special reference to political 
and economic action. In this view both formal institutional patterns 
and informal cultural norms, in and of themselves, are not regarded 
as very reliable guides to a nation's "character." Those adopting this 
approach stress particularly the history of peoples or societies, and 
on this basis may characterize them as warlike or peaceful, enter- 
prising or backward, trustworthy or deceptive, pragmatic and 
industrious, or idealistic and impractical. Germany is a case often 
discussed in this context. Many have emphasized the contrast be- 
tween Germany's outstanding institutional creations and cultural 
achievements on the one hand, and on the other its historic role in 
Europe in the first half of the twentieth century. Hearnshaw's Ger- 
many the Aggressor Throughout the Ages (1940) may serve as an 
example. This mode of analysis should not be confused with a more 
sophisticated type in which national character is recognized to be a 
property of persons, and is treated as an independent variable con- 
tributing to an explanation of some form of political action con- 
sidered as a dependent variable. An outstanding example is Gabriel 
Almond's (1950) use of materials on the American character to ex- 
plain certain persistent tendencies in the conduct of foreign policy 
by the United States. 


National Character as Racial Psychology. The identification 
of national character with the allegedly "inborn" and presumably 
biological characteristics (generally defined as superior or inferior) 
of a group is one of the oldest and most common approaches, and in 
modern social science the one most severely criticized if not actively 
abhorred (cf. Benedict 1945) . A typical illustration, by no means 
the most extreme, may be found in Jaensch's (1938) study, pub- 
lished under Hitler, in which he asserted that the French were usu- 
ally erratic and unreliable, the Germans consistent and stable. 

The belief in racial psychology is by no means restricted to racist 
theoreticians. As tolerant and democratic a man as Andre Siegfried 
(195 1), for example, attributes one of the two main qualities he 
finds in the French mind — its being "extremely practical and mat- 
ter of fact" — to a Celtic heritage which he says is found wherever 
"Celtic blood prevails," including places as widely separated as 
northern Spain and the west of the British Isles. And Brickner's 
( 1943 ) analysis of the German character as one essentially paranoid 
struck many students of the problem as verging on racism in psy- 
chology, even though it certainly did not suggest that the allegedly 
typical paranoid behavior was biological in origin. Although the 
pendulum may have swung too far in the opposite direction, there 
is today general agreement that the biologically given properties 
of what are in any event extraordinarily mixed national populations 
are not a significant influence in shaping the institutions, culture, 
or behavior of those national populations. Yet the altogether proper 
discrediting of racial psychology has perhaps had the unfortunate 
unintended effect of discouraging serious scientific research on a 
basic question of social science. 

In most of the better known general essays on national character, 
such as those by Sforza (1942) on Italy, Siegfried (1930) on 
France, and Ortega y Gasset (1937) on Spain, more than one of 
these definitions or approaches will be used simultaneously and gen- 
erally without any special note being taken of this fact. Typically, 
no distinction is made between character as something already 
formed and acting, and those forces such as climate and geography, 
history, biology, or child rearing which may be designated as the 
causes or consequences of the observed national character. If prog- 
ress is to be made in the field, we need to make our investigations 
more systematic. There is no one line of development which can do 
full justice to the complexities of the problem. We feel, however, 
that great advantages inhere in the concentration on modal adult 


personality characteristics as a central problem in national charac- 
ter study. We therefore pose the question: whether produced by 
common heritage, common upbringing, the sharing of common 
culture, the exposure to common institutional pressures, or other 
causes, are there in fact any clearly demonstrated important diflfer- 
ences in the psychological characteristics of the populations who 
make up modern national states? The question is more difficult to 
answer with confidence than many imagine it to be. 

The Problem of Measurement 

No matter how we conceive of national character, a scientific 
approach to it must face the problem of its assessment — or to use a 
less evasive word, its measurement. This subject generates as much 
confusion and malaise as does the issue of definition. The different 
approaches to national character based on institutional structure, 
and on national action or behavior, involve virtually no common 
understanding, standard techniques, regular procedures, or canons 
of reliability and validity. The situation is only slightly less variable 
in the racial psychology and the culture-pattern approaches. Each 
study proceeds almost entirely independently of all others, utilizes 
unique perspectives, draws on distinctive materials, follows idiosyn- 
cratic rules of evidence, and observes only its own standards of re- 
liability and validity. The result is, if not intellectual chaos or 
anarchy, at least a great buzzing, blooming confusion which defies 
representation. Under the circumstances, a systematic comparative 
perspective is almost impossible. 

It is argued by some, not without cogency, that institutional ar- 
rangements are so varied, culture patterns so unique, national psy- 
chologies so distinctive, that no common or standard language can 
hope to encompass this infinite diversity. Under these circum- 
stances, it is said, we cannot do justice to the unique character of any 
people unless we develop a special battery of concepts and a new 
glossary of terms to describe them. This claim may be somewhat 
exaggerated. In any event it suggests that systematic analysis of 
national character as a field of scientific investigation is blocked. 
The same basic difficulty does not, at least in equal degree, attend 
efforts to deal with national character as modal personality patterns. 
There is good reason to believe that the range of variation in human 
personality, however great, can be adequately encompassed by a 
conceptual scheme, with a sufficiently limited set of terms to make 
for manageable research designs without sacrifice of essential rich- 



ness or variety. We also maintain that, despite the many methodo- 
logical and conceptual problems involved, this scheme and its meas- 
uring instruments can be developed so as to permit reliable and valid 
applications across national lines. 

Harold Lasswell once claimed it would be an exaggeration to say 
that in two thousand years of studying politics we had made no ad- 
vances whatsoever beyond Plato and Aristotle. Perhaps an exaggera- 
tion, but not a great one. At least so it seems when we recognize 
that the genius of political analysis has gone mainly into the inven- 
tion of new terms for old ideas which were never made operational, 
never tested, and therefore never developed. For how else is one to 
choose between Plato's theory of the desiring, spirited, and reason- 
ing parts, Pareto's "residues of combination" and "residues of per- 
sistence of aggregates," Spranger's six types of men, or Thomas and 
Znaniecki's Philistine, Bohemian, and Creative Man. These ap- 
proaches must meet the criticism, as Spranger acknowledged, that 
they "abandon the concrete ground of experience and reduce 
psychology to mere speculation" ( 1928 :xi) . 

As Harold Lasswell went on to say, however, our chief contem- 
porary advantage over Plato and Aristotle lies "in the invention and 
adaptation of procedures by which specific individuals and groups, 
operating in specific historic and cultural settings, can be under- 
stood. ... In a word, the modern approach is toward the building 
of scientific knowledge by perfecting the instrumentalities of in- 
quiry" (1951:468-469). For the first time in the history of the 
study of politics we actually have within our grasp the means for 
systematic study of such conceptions as those developed by Plato, 
Pareto, and Spranger. I refer, of course, to the great strides made 
in this century in our understanding of personality dynamics and 
in the means for personality testing, measurement, and assessment. 
However, the concepts of Plato and others must first be clarified. 
They must be made operational, that is, transformed into possible 
research procedures of testing and measurement. 

In some cases this has already been attempted, and it has been 
found possible and useful to devise formal measures of these classic 
typologies. Spranger's types, for example, were an important in- 
fluence in shaping the widely used Allport- Vernon Scale of Values. 
In the process the old concepts may be found wanting. For exam- 
ple, Lurie's (1937) factor analysis to ascertain which generalized 
attitude clusters, if any, conform to Spranger's types, located sev- 
eral fitting Spranger's definition fairly closely — the theoretical, the 


religious, the social, and the economic-political. Several others, how- 
ever, could not be empirically distinguished. As we test and perhaps 
discard some of these "classic" concepts, they will be replaced by 
others which are proving important in our study of personality and 
have obvious relevance to politics, such as: the needs for power, af- 
filiation, and achievement; the authoritarian and ethnocentric syn- 
drome; dominance drives; alienation and anomie; dogmatism and 
rigidity; tough- and tender-mindedness. It is in the nature of sci- 
ence and the inevitable path of its advance that concepts are re- 
placed as empirical research advances. If for sentimental reasons we 
are unable to abandon the old familiar concepts, we may do our- 
selves honor as classicists, but we disqualify ourselves as scientists. 


The definition and classification of political systems is a more 
familiar and less ambiguous task, although it too has its vicissitudes. 
The sturdy old distinctions among political forms such as democ- 
racy, oligarchy, and tyranny which come down from Plato and 
Aristotle still serve us well today, although some may prefer a more 
contemporary classification, such as that proposed by Gabriel 
Almond (1956) who identifies the Anglo-American, the Con- 
tinental European, the pre- or partially industrial, and the totali- 
tarian political systems. Whatever scheme we might choose, we 
would probably not have great difficulty in agreeing on the defining 
characteristics of each type and could probably attain fair agree- 
ment in classifying particular societies. 

Such classifications are, however, deceptively easy, and for many 
purposes they may be misleading. We generally accept the Greek 
city-state as the epitome of the democratic political system, but we 
should not forget that internally it rested squarely on a large slave 
class, and in external affairs was characterized by almost continuous 
intercity warfare motivated by nothing more noble than the desire 
for power and gain. Tsarist Russia was perhaps the most absolute 
autocracy in Europe in the eighteenth and nineteenth centuries, yet 
the village 7nir was a self-governing community observing some of 
the purest principles of egalitarian democracy. Germany was an 
outstanding example of relatively absolute monarchy before World 
War I, although intellectually and spiritually one of the freest na- 
tions in Europe. The Weimar Republic which followed represented 
the embodiment of the most advanced democratic principles, but it 
was succeeded by one of the blackest of totalitarian regimes — which 


again is followed by a West German Republic which seems one of 
the stablest and most genuine of Europe's democracies. The rule of 
Ataturk in Turkey was a dictatorship, yet he used his dictatorial 
powers to foster democratic institutions against the resistance of 
the traditional religious oligarchy and the peasant masses. Soviet 
Russia under Stalin had what was nominally the most democratic 
constitution in the world, while in fact it closely approximated a 
regime of absolute totalitarian terror. 

The obvious point is that we must differentiate the components 
of political systems just as we must distinguish the diverse elements 
in, and the different bearers of, national character. As a minimum 
we must make a distinction between: the relatively enduring and 
the more fleeting or transitional features of a nation's political sys- 
tem (cf. Lipset i960 on stable and unstable democracies) ; the 
formal, exoteric system from the informal, esoteric, operational 
patterns (cf. Leites 195 1 on the Politburo) ; the politics of central 
government from that which characterizes vital institutions such 
as the local community, the church, trade union, or family (cf. 
Michels 1949 on the iron law of oligarchy) ; the principles embodied 
in constitutions and other venerated documents and those com- 
monly held by the populace (cf. Stouffer 1955 on civil liberties in 
the United States) ; the political orientation of the elite as against 
that of the rank and file of the population (cf. Stouffer 1955 and 
Mills 1956 on the power elite). 

Only if we recognize both politics and national character as 
highly differentiated systems of variables can we hope to do any 
justice to the complex phenomena we are studying. Unfortunately 
many, indeed most, studies which seek to relate character to politi- 
cal systems fail to make these necessary distinctions. They treat 
political systems as undifferentiated and more or less unchanging 
units rather than as complex variables. ^ 


Despite the efflorescence of the field of culture and personality 
during the last three decades,^ and a parallel growth of interest in 

The point at which a new field of exploration begins can as a rule be designated only on 
an essentially arbitrary basis. Most authorities acknowledge Franz Boas as the father of this move- 
ment (see especially Boas 19 lo), and many date its formal beginning with the publication in 1934 of 
Ruth Benedict's Patterns of Culture. Ruth Benedict and Margaret Mead were, of course, students in 
the seminars on Individual and Society which Boas gave at Columbia in the late twenties. Boas 
himself gave great credit to Theodore Waitz, of whose Anthropolgie der Naiurvolker he said 
"[this] great work is an inquiry into whether there are any fundamental differences between the 
mental make-up of mankind the world over, racially as well as socially." 


the empirical study of modern political systems, we can point to 
very few systematic empirical studies of the relations between per- 
sonality patterns, or psychological factors in general, and the rise, 
functioning, and change of political systems. As usual the history of 
intellectual disciplines reveals much of the story. Modern studies 
of the relations between personality and sociocultural systems have 
been developed almost exclusively by cultural anthropologists. Per- 
haps because most nonliterate (or primitive) people rarely have a 
formal or specialized political organization, all but a few cultural 
anthropologists have shown little interest in political structure. In 
this respect, at least, the students of personality and culture have 
followed the dominant pattern in their discipline. Benedict's book 
on Japan (1946) and Hsu's comparison of the Chinese and Ameri- 
can culture ( 1953 ) each give a chapter or more to politics and gov- 
ernment, and Mead (1957) devoted an entire book to Soviet atti- 
tudes toward authority, particularly political authority. But these 
are outstanding exceptions. The early editions of the two standard 
and massive American collections of articles on culture and person- 
ality do not contain a single item which deals directly with the rela- 
tion of personality patterns to the political system.^ Similarly, the 
standard anthropological textbook in the field contains a chapter 
on psychiatric disorders and one on "personality in class, caste, re- 
gion, and occupation," but none on politics.^ Linton's (1945) lit- 
tle classic on The Cultural Background of Personality makes no 
mention of government or politics. The same may be said of the 
works of Abram Kardiner (1939, 1945) which have done so much 
to shape the field. Geoffrey Gorer's study of the English character 
has chapters on "friends and neighbors," on "people and homes," 
on "religion," and on "marriage," but none on those political in- 

° Clyde Kluckhohn and Henry Murray (1953); Douglas Haring (1948). The former did 
contain an article on personality under the Nazis, but rather than having a political focus it 
was designed only to show that personality remained unchanged despite changes in the indi- 
vidual's political security. The latter had an article on the armaments race, but only as illus- 
trating a type of mechanism in interpersonal relations. Later editions gave somewhat, but not 
much more, attention to the political process. The later edition of the Kluckhohn, Murray (and 
Schneider) volume (1956) included a new article by R. Bauer, "Psychology of the Soviet 
Middle Elite." In addition, the third edition of the Haring volume (1956) included materials 
on the role of character in postwar Japanese sociopolitical development and one by Gorer 
which, while not explicitly dealing with political structure, discussed the role of the police in 
the apparent modification of the English character in modern times. 

'John Honigmann (1954). The index does call attention, under the heading "political re- 
lations," to two pages which discuss the evidence that organizational atomism in a community 
is related to the degree of ingroup sorcery, and two pages on the relations of family patterns 
to political structure. 


stitutions and attitudes about parliaments, elections, local govern- 
ment, civil liberties, and personal rights which most people regard 
as the truly distinctive political features of English society/ 

These comments are, of course, not meant to ignore the substan- 
tial contribution of the British anthropologists to our understand- 
ing of primitive political systems, but in this case the hiatus is com- 
plementary to that found in the culture and personality studies. 
In their exceptionally fine work on African political systems Fortes, 
Evans-Pritchard, and their associates (1940) say virtually nothing 
about the characterological qualities which may be important to 
the development and maintenance of stable political orders in these 
important underdeveloped regions. 

Unfortunately the situation is not markedly changed when we 
consider the work of political scientists, to whom one might ap- 
propriately assign greater responsibility for this line of work. 
Although Plato and Aristotle both stressed the role of character in 
shaping political forms and processes, the person tends periodically 
to disappear from political theory. Early in this century Graham 
Wallas made a plea for a return to the study of human nature in 
politics. He deplored the books by American university professors 
as useless, because the writers "dealt with abstract men, formed on 
assumptions of which they were unaware and which they had never 
tested either by experience or by study" (1908 : 10) . Very little was 
done to take up the challenge. More than two decades later Charles 
E. Merriam (1925) was still pleading the same needs, but in a more 
focused and hopeful manner with emphasis on personality, meas- 
urement, large-scale statistical studies, and correlational analysis 
of the relations between political conduct and psychological char- 
acteristics of the political man. In the same year Henry Moore 
(1925) published a pioneering study of psychological factors as- 
sociated with holding radical and conservative political opinions. 
Moore's analysis, utilizing tests for resistance to majority opinion 
and of readiness to break old habits, anticipated much of the recent 
research on personality and politics. Unfortunately it failed to be- 
come the start of an active research tradition in psychology. 

Merriam's role in fostering the application of psychology to 
politics is comparable to that played by Franz Boas in the develop- 
ment of culture and personality studies. It was under Merriam's 

* Gorer's (1955) book does contain a chapter on "law and order," but it deals exclusively 
with two questions: the popular image of the police and the attitude toward "fiddling," a 
term used to describe minor infractions of the rationing regulations. 


influence that Harold Lasswell wrote what was probably the first 
modern, systematic, and broad application of psychology to con- 
temporary politics. In Tsycho pathology and Politics (1930) Lass- 
well broke new ground in going beyond the usual hypothetical 
classification of political types to develop the detailed study of life 
histories. Guided by psychoanalytic theory, he showed quite ex- 
plictly and empirically the connection between personality traits 
and the choice and style of political roles such as the agitator, the 
propagandist, and the administrator. In the same volume he 
sketched one of the first systematic schemes for describing per- 
sonality in politically relevant terms. Although he worked mainly 
with the individual case study, Lasswell was not unaware of the im- 
plications of this mode of analysis for the study of political pat- 
terns characteristic of classes and national populations. "What mat- 
ters to the student of culture," he said, "is not the subjective 
similarities of the species but the subjective differences among the 
members of the same and similar cultures" (1930:261). He did not, 
however, follow through to undertake the systematic research this 
statement implied. 

A decade elapsed before the next really major event in the field 
occurred with the publication of Erich Fromm's Escape from Free- 
dom (1941). Fromm took the step that Lasswell had anticipated 
but failed to make himself. He held that the typical character types 
prevalent at any given time were different, that these differences 
varied systematically with changes in the socioeconomic system, and 
that character types could serve either as a cement holding the sys- 
tem together or as an explosive tearing it apart, depending on the 
degree to which a given character type fit the demands of the sys- 
tem and found satisfaction in it. He traced this interaction through 
the history of medieval Europe and the Reformation, sought to ex- 
plain the appeal of Hitler by the widespread prevalence of the au- 
thoritarian character in Germany, and sketched some of the forces 
in democratic society — such as the sense of aloneness, the loss of 
individuality and spontaneity — which he saw as inducing an 
"escape from freedom." 

Fromm's theory has been extraordinarily stimulating to all con- 
cerned with the study of personality and politics. We should appre- 
ciate his theoretical sophistication, his clinical intuition, and his 
clear recognition of the most vital problems. His use of historical 
documents and contemporary sources, such as political speeches 
and party platforms, represented a commendable improvement 


over the efforts of those who were content to rely more or less ex- 
clusively on their clinical experience with psychoanalytic patients. 
Nevertheless, many students of the problem would insist that 
Fromm's analysis did not present more than suggestive hypotheses. 
It was yet to be demonstrated by objectively verified testing based 
on adequate samples that the modal personality types in different 
socioeconomic systems were significantly different from each other, 
or that within any nation the form and content of political action 
varied according to the personality traits typical for any group. 

Considering that the conflict of political principles played so 
central a role among the issues in World War II, it is rather striking 
that the series of books on national character which anthropologists 
contributed to the war effort gave such incidental, indeed almost 
casual, treatment to the relations between national character and 
democratic government. There are important limitations on the 
justice with which this characterization can be applied in one or 
another case, yet it fairly well fits the work of Gorer on Japan 
(1943) , Russia (1950), and the United States (1948) , Mead on the 
United States (1942), and Benedict on Japan (1946). Insofar as 
they did deal with governments, they did not with any rigor specify 
the personality traits of politically active adults which might con- 
duce them to support democratic or autocratic government. In- 
stead, their method was to highlight the analogy between the po- 
litical system and other features of the culture, most notably the 
family. Thus Gorer notes the characteristic division of power in 
the United States as contrasted with greater centralization in Euro- 
pean governments, then points to the typical American nuclear 
family council, and concludes that "to a certain extent the pattern 
of authority in the state is reflected in the family" ( 1948 : 44— 45) . 
Similarly, Benedict notes that the Japanese father is not a martinet, 
but rather exercises his authority as the representative of the larger 
family. The attitude thus "learned by the child in his earliest ex- 
periences with his father" is then invoked to explain why in Japa- 
nese governmental affairs "the officials who head the hierarchy do 
not typically exercize the actual authority" (1946:301). 

These are undoubtedly important insights. Nevertheless, to con- 
ceive of the family as the mirror of the state, and of the state as a 
reflection of the pattern of relations in the family, establishes a 
circle without any suggestion as to how change can and does come 
about. In the case of the Japanese, Benedict sought to meet this 
challenge by stressing the Japanese "ethic of alternatives." But 


what of the Germans and Russians who presumably do not have 
such an ethic? Are they doomed to perpetual authoritarian govern- 
ment as the cycles of family and state patterns ever renew them- 

The basic difficulty with this approach, one pervasive in the cul- 
ture and personality literature, is its failure to take adequate ac- 
count of the differentiation within large national populations. It 
emphasizes the central tendency, the existence of which it presumes 
but does not prove, and neglects the range of variation within and 
around the average or typical. Once we begin to deal with distri- 
butions, with variation and range, we must recognize that a second 
weakness of this approach is that its descriptive language, the tech- 
nical terms on which it is based, does not easily permit the precise 
measurement and quantitative expression necessary to the study of 
a distributive phenomenon. These deficiencies were largely reme- 
died in another set of the wartime studies, particularly those by 
Henry Dicks (1950) and David Levy ( 1 951) , which represent an 
important landmark in the development of our understanding of 
how personality relates to political action. 

Dr. Dicks' work was in the main line of culture and personality 
studies in that it considered personality in psychoanalytic terms and 
was based on a general model of the German personality drawn from 
a variety of cultural sources. In his case, however, what is generally 
the conclusion of many studies was only the starting point. He went 
beyond previous studies in three important respects: (i) the per- 
sonality of each subject was explicitly scored on clearly specified 
andcarefully defined variables; (2) the political orientation of each 
person was also carefully measured in concrete terms; and (3) the 
personality measures and the indices of political orientation were 
systematically related to each other by standard statistical pro- 
cedures. All this was done with clinical sensitivity, with use of gen- 
eral theory, and without loss of contact with the more traditional 
but impressionistic description of the German national character. 

Dicks worked with a sample of 138 German soldiers taken as 
prisoners of war between 1942 and 1944. On the basis of politically 
focused interviews each man was classified on a five-point scale 
running from "fanatical, wholehearted Nazi" to "active, convinced 
anti-Nazi." In addition, on the basis of nominally free but in fact 
highly focused psychiatrically oriented interviews, each man was 
rated on 1 5 different psychosociological variables ranging from de- 
gree of religiosity to presence or absence of schizoid features. Re- 


lationships attaining a high degree of statistical significance (at the 
.01 level or better) were obtained between Nazism and six of the 
fifteen psychosocial variables. For example, those high on the scale 
of Naziism showed a marked taboo against tenderness, were more 
sadistic or antisocial, and were much more likely to engage in pro- 

It is important to recognize that Dicks did not prove these or 
any other characteristics to be generally present in German na- 
tionals. He proved only that Nazis and near-Nazis were different 
from non-Nazi Germans in a number of important respects. This 
is not to say that Dicks did not attempt a general characterization 
of the German personality. He could hardly have undertaken his 
study without some such hypothetical model which, he assumed, the 
Nazi "embodied in more exaggerated or concentrated form." The 
typical German he described as having "an ambivalent, compulsive 
character structure with the emphasis on submissive dominant con- 
formity, a strong counter-cathexis of the virtues of duty, of 'con- 
trol' by the self, especially buttressed by re-projected 'external' 
super-ego symbols." Even though such individuals might be highly 
•susceptible to the propaganda themes and the style of leadership 
offered by the Nazis, it is also apparent that this character type 
could freely support any one of a number of different sociopolitical 
orders. Dr. Dicks' study is of particular value, therefore, in keeping 
before us the awareness that in any national population there is 
likely to be substantial variation in modal personality patterns, even 
though for any given nation this variation may cover only a narrow 
part of the world-wide range. Dicks' study also suggests that the 
extreme political positions are those which are most likely to be 
attractive to the extremes on the personality continuum. If the ex- 
tremists seize power, the resulting political forms may or may not 
be congruent with the dominant personality tendencies in the popu- 
lation at large. It seems likely that this congruence was greater in 
Hitlerite Germany than in Stalinist Russia. 

Inkeles, Hanfmann, and Beier (1958) administered a battery of 
tests including the Rorschach, TAT, sentence-completion test, and 
others to a small sample (51 cases) of refugees from Soviet Russia 
who departed during and just after World War II ( cf. Dicks 1952). 
On this basis they constructed a composite national character por- 
trait, differentiating a main modal pattern, a variant on it, and a 
residual group. The subjects were also divided into four social 
classes. The authors did not, unfortunately, relate the personality 


characteristics of each individual directly to his mode of political 
orientation. For the group as a whole, however, they related its 
adjustment to the Soviet political system to each element of the 
modal personality pattern — which included a strong need for af- 
filiation, marked dependency needs, emotional expressiveness and 
responsiveness, and resistance to being shamed for failures in imper- 
sonal performance. The authors found, for example, that the per- 
sistent shortages of food, shelter, and clothing which characterized 
Soviet life under Stalin, aggravated the anxieties about oral depri- 
vation which were frequently manifested in the Russian character. 
In general, they concluded, "there was a high degree of incongru- 
ence between the central personality modes and dispositions of 
many Russians and . . . the behavior of the regime." This was most 
marked, however, for those who represented the basic personality 
mode, and was much less true for those whose personality reflected 
a substantial departure from the modal pattern common to the 
mass of peasants and workers. 

Postwar Developments 

Research in the period after World War II has been characterized 
by two important developments : ( i ) improvements in the methods 
for assessing personality on a large scale and (2) the application of 
such methods on a cross-national or comparative basis. 

If we require that national character studies be based on syste- 
matic and objective study of personality, that they represent all the 
diverse elements of national populations, and that they permit 
meaningful comparison with results from other studies, we are in 
effect calling for a transformation of the standard methodology of 
the field. Such a demand made before 1940 would have been perhaps 
not visionary, but hardly reasonable as a practical matter. The post- 
war period, however, has seen the development and application of 
means for the assessment of personality which enable us to measure 
it with relative ease, and to do so with large representative samples. 
There is reason to believe that at least some of these instruments 
may be effectively used cross-nationally. 

The effort to measure with some precision the personality traits 
of entire national groups has a longer history than many suppose. 
One of the earliest ventures in the use of a standard psychological 
test to assess personality trends in a significantly large population 
was the Bleulers' (1935) application of the Rorschach Ink-Blot test 
to Moroccans in the thirties. The Bleulers administered the Rors- 
chach to an unspecified number of "simple country folk" (half 


Arab, half Berber) living in the vast plains of West Morocco. Their 
characterization, based on the Rorschach records as measured 
against their experience with the test in Europe, is full of comments 
of the following order: the Moroccan lacks the typical European 
"tendency to abstractive generalization;" his extroversion emerges 
mainly in "a marked enthusiasm under the influence of momentary 
events . . . but he lacks the systematic, energetic, and persevering 
striving after outward success." 

Of course we will wonder whether we can safely generalize these 
comments to other Moroccans, and how much these patterns reflect 
not Moroccan culture but rather the low level of education and the 
relative isolation of these people. But more important for our pur- 
poses is the question of the relevance of such qualities of character 
for the ability to act as a good citizen in a stable political order of a 
national state. The Bleulers' description typically makes no mention 
of images of authority, civic consciousness, or other traits of obvi- 
ous political relevance, and we do not have the knowledge to judge 
whether the lack of a tendency to abstractive generalization is con- 
ducive to good democratic citizenship or not. That these defects 
of the typical Rorschach analysis of group personality are relatively 
persistent may be observed by comparing the Bleulers' study with 
later ventures, such as the study of the Chinese by Abel and Hsu 
(1949) . Indeed, the Rorschach has come into serious question as an 
instrument for systematic research into group traits (Carstairs, 
Payne, and Whitaker i960) . 

Probably the greatest influence on our thinking and practice in 
the measurement of personality dimensions relevant to politics is 
exerted by the now classic study of the authoritarian personality by 
the Frankfurt Institut fiir Sozialforschung (Horkheimer 1936). 
Erich Fromm played a major role in this group's development of the 
concept of the authoritarian personality, which Adorno (1950) 
and his associates carried forward in the United States both theo- 
retically and methodologically. The main fruit of the California 
group's investigation was the isolation, definition, and measurement 
of a particular personality type, but the conception of that type 
was initially derived from ideas about the distinctive psychological 
coloration of authoritarian political creeds and movements. 
Although the F scale ^ has been severely criticized because it can 

The letter F was used with the scale to designate "susceptibility to Fascism." This sounds 
more like a specifically political than a psychological measure, although the authors intended it 
mainly as a measure of personality. This use of the term Fascism for the scale unfortunately 
clouded the issue by seeming to prejudge the relation between measures of personality and those 
of political orientation, or worse to suggest they were perhaps one and the same thing. 


distinguish right authoritarians but permits left authoritarians to 
escape notice (Christie 1954) , there can be no serious question but 
that the psychological syndrome thus isolated is highly correlated 
with extreme right-wing political attitudes. 

The semipsychiatric interview which Dicks used requires special 
talent to conduct, is difficult and expensive to code or score, and 
must therefore be restricted to very small samples. By contrast the 
F scale has the special virtue of great simplicity as a test instrument, 
something unusual in the earlier efforts to measure personality 
variables of theoretical interest and proved clinical significance. 
The F scale thus made possible for the first time the simultaneous 
collection of data on personality and on political orientations from 
a fully representative national sample. Using a modified version of 
the F scale, Janowitz and Marvick found that in the United States 
those whose personality tended more toward authoritarianism were 
also more markedly isolationist in foreign affairs (cf. Levinson 
1957) . The more authoritarian also revealed a sense of political in- 
effectiveness, that is, they believed themselves powerless to influence 
government action. The conclusion reached by Janowitz and Mar- 
vick is particularly noteworthy: ''Personality tendencies measured 
by [an] authoritarian scale served to explain political behavior at 
least as well as those factors [such as age, education, and class] tra- 
ditionally included in political and voting behavior studies." ( 1953 : 
201; also see Lane 1955.) 

In addition to the F scale, there are other personality measures 
suitable for administration to large samples and relevant to political 
orientations, such as Rokeach's (1956) dogmatism scale and 
Eysenck's (1954) classification of the tender minded and tough 
minded. In their study of American automobile workers, Arthur 
Kornhauser (1956) and his associates utilized measures not only of 
authoritarianism but also of life satisfaction and social alienation or 
"anomie." Those characterized by anomie showed little interest in 
politics, and were much less likely to vote. When they did vote, they 
tended to vote contrary to the prevailing sentiment among their fel- 
low workers. Among numerous important findings in this rich and 
interesting study was the discovery that authoritarianism is re- 
lated to political extremism ivhether of the right or left. This as- 
sumption gains support from a study of political orientations in 
Iran. Despite their fundamental differences in political position, the 
extreme rightists and extreme leftists were more like each other in 
many social and behavioral characteristics — such as "level of social 


detachment" and "breadth of social horizons" — than they were 
Hke the more moderate groups of the pohtical center (Ringer and 
Sills 1953). 

In summarizing their detailed results, Kornhauser and his asso- 
ciates reach a conclusion which accords well with the requirements 
of our model of the democratic personality. They say: "The prob- 
lem of democracy ... is partly the problem of maintaining an 
adequate proportion of members who are capable of engaging in 
the market place of proposals and counter-proposals, immune from 
the feeling that 'the leader knows best' and from the temptation 
to condone, or to resort to, desperate measures in times of social 
and political crisis" (1956:249-250) . 

Perhaps the most systematic effort to relate personality to politi- 
cal inclinations is to be found in the pioneering study by Herbert 
McClosky (1953) in which he sought to define the personality 
characteristics of those taking positions along the continuum from 
conservative to liberal politics. He unfortunately defines conserva- 
tism not by party affiliation, but on the basis of agreement with a 
set of normative propositions drawn from the works of leading, 
modern, conservative spokesmen. These statements include items 
such as: you can't change human nature; no matter what people 
think, a few people will always run things anway; duties are more 
important than rights. Using a rich battery of personality scales 
developed at the University of Minnesota and elsewhere, he finds 
that the extreme conservatives are sharply differentiated from both 
the "liberals" and "moderate liberals" in being more submissive, 
anomic, alienated, pessimistic, guilty, hostile, rigid, paranoid obses- 
sive, intolerant of human frailty, and extremely ego-defensive. It 
will be immediately apparent that the personality traits of the ex- 
treme conservative or "reactionary" bear a very close relation to 
those of the authoritarian personality, and at every point are polar 
to the qualities described below in our model of the democratic per- 

It is unfortunately characteristic of McClosky 's study, and many 
others in this field, that they are not comparative. This necessarily 
leaves us in doubt as to whether in other countries or environments 
the same traits of personality would also be associated with the same 
kinds of political orientation. For example, Dr. Dicks' (1950) 
study raises at once a question as to the uniqueness of the Nazi pat- 
tern and the degree to which we can generalize his findings. Since 
all of Dicks' comparisons were made within the German sample, 


he is quite justified in saying that in Germany certain individual 
characteristics are more associated with fascist poHtical leanings 
than others. But his assumption that the Nazis are only extreme 
variants of a more general or typical German character cannot be 
taken as proved. On the basis of his sample he could hardly estab- 
lish what the average or typical German is like, if he exists at all. 
In any study restricted to one sample, we may easily be led into 
assuming that the response which fits our preconception of the 
group is distinctive to it, when in fact that response is quite com- 
mon in other populations as well. For example, we would have 
much more confidence in Schaffner's (1948) finding of extreme 
authoritarianism in the typical German conception of the family 
had he given his sentence completion test to at least one other com- 
parable national group. 

This defect was remedied in a number of studies conducted after 
World War II. Indeed the postwar period is outstanding for the 
development of more systematic comparative research. For example, 
D. V. McGranahan (1946) put a number of questions on basic 
issues — such as obedience to authority under duress, and freedom 
of the press even when not "for the good of the people" — to com- 
parable samples of American and German boys. In the latter case 
he made a distinction by political orientation between Nazis, neu- 
trals, and anti-Nazis. The German youth distinctly favored obedi- 
ence to authority more often than the Americans, showed less faith 
in the common man, and were more admiring of people with po- 
litical or military power. In general these findings fit our expecta- 
tion with regard to the greater emphasis on democratic values in 
American as against German society. But it is crucial to note that 
within the German group, those classified as anti-Nazi were on 
some questions closer to the Americans than to their Nazi-oriented 

Of course, no simple conclusions can be drawn from one such 
study standing alone. For example, when the same questions were 
given by Stoodley (1957) to a more or less comparable group of 
youths from the Philippines, he found that on some dimensions 
they were closer to the Germans, on others, to the Americans, thus 
yielding a distinctive national profile. Unfortunately, he did not 
inquire into the relation of these attitudes to political orientation, 
which would have enabled us to judge whether the same value 
orientations which made for Nazism in Germany made for com- 
parable antidemocratic leanings in the Philippines. 


Gillespie and Allport ( 1955 ) studied hopes for the future among 
college students in several countries. Although they did not inquire 
directly into political beliefs, several of the topics they dealt with 
are clearly relevant to an evaluation of the strength of tendencies 
toward various forms of active "citizenship." They reported the 
Japanese to be outstanding in their "sense of obligation to the social 
group in which they live." The Japanese were, for example, first 
among all countries in saying they would seek to inculcate in their 
children such qualities as good citizenship, social usefulness, and 
service to society {cf. Stoetzel 1955) . On this and similar questions 
Americans were near the bottom of the list. They "emphasized 
their rights rather than their duties and in all presented a picture 
of individuality, separation from the social context of living, and 
privatization of values and personal plans" (1955:29). The New 
Zealanders presented a profile quite similar to that of the Ameri- 
cans, but we cannot say whether this results from their common 
Anglo-Saxon heritage, the common experience of setthng a new 
continent, or some combination of these and similar influences. 
These findings are well in accord with the conclusions of earlier, 
more impressionistic studies of American and Japanese character. 
They are none the less welcome for providing firm confirmation of 
these hypotheses. 

Despite such promising starts there seems to be great hesitation 
to undertake systematic comparative studies. The hesitation to 
apply methods of personality testing cross-nationally arises not 
merely from the magnitude and cost of the task, admittedly sub- 
stantial, but in large part from resistance, skepticism, and outright 
rejection of the possibility of reliable and valid cross-national test- 
ing of opinions, values, and personality traits. We should not mini- 
mize the substantial technical difficulties facing any such effort. But 
the objections often offered to such attempts seem exaggerated, and 
in any event the appropriate response is to accept the challenge 
and attempt the necessary methodological innovation. By way of 
encouragement we may note that a number of studies have shown 
that certain tests can be used cross-nationally with a high degree of 
reliability. In a study for UNESCO (Cantril and Buchanan 1953) 
conducted in nine countries it was found that most questions had 
the same meaning in all the countries studied, and that the opin- 
ions related to each other in one setting were similarly correlated in 
the others. For example, in each country those who believed human 
nature can be changed were also more likely to believe that national 


characteristics arise from the way in which people are brought up. 
Indeed the same syndrome, or complex pattern of attitudes, was 
represented in all countries. One group in each country, who might 
be called the optimists, believed human nature perfectible, national 
character pliable, world peace attainable, and world organization 
desirable. The pessimists, or fatalists, believed there would always 
be wars, human nature cannot be changed, and that efforts at im- 
proving the international situation are bound to fail. 

The UNESCO study, of course, dealt more with opinions than 
with deeper lying attitudes and facets of personality, but we are not 
limited to that level. In an important study of values which Charles 
Morris (1956) conducted in the United States, India, and China, 
he discovered that in each country the ratings of individual ques- 
tions were made along the same common value dimensions and 
that "there is thus revealed an underlying value structure (or value 
space) which is very much the same in the culturally diverse groups 
of students." In addition, the relation of the value factors to other 
issues was much the same in each culturally distinct group. For 
example, those individuals whose values centered on receptivity to 
and sympathetic concern for others tended, in all three countries, to 
dislike or reject the operative values of the political world, as meas- 
ured by the Allport-Vernon scale. 

Similar results are reported in the use of a personality test which 
presumably taps deeper-lying strata of the personality. In a com- 
parative study of teachers in seven European countries it was found 
that the same items of the F scale designed to test authoritarianism 
tended to cohere and form a pattern in all of the countries studied.*' 
In addition, the research uncovered high consistency in the way in 
which orientations toward threatening situations in both domestic 
and international politics were patterned in the several countries. 
But at the same time the authors offer us some sobering words of 
caution regarding the difficulties facing such comparative studies. 
They found "many of the relationships vary in size, direction, and 
significance in different countries . . . modified by specific national 
and international situational factors — by the historically given 
structures of political forces, by the dominant policies, by majority- 
minority relations, by the ongoing communication processes in the 
mass media and in the larger organizations" (Aubert et al. 1954: 

'' Personal communication from Drs. D. J. Levinson and Stem Rokkan. The data were col- 
lected in the study reported in Aubert, 1954. 



It is apparent that we have made at least a modest beginning in 
studying the relation of personality patterns to the development 
and maintenance of political systems. There is substantial and rather 
compelling evidence of a regular and intimate connection between 
personality and the mode of political participation by individuals 
and groups within any one political system. In many different in- 
stitutional settings and in many parts of the world, those who ad- 
here to the more extreme political positions have distinctive per- 
sonality traits separating them from those taking more moderate 
positions in the same setting. The formal or explicit "content" of 
one's political orientation — left or right, conservative or radical, 
pro- or antilabor — may be determined mainly by more "extrinsic" 
characteristics such as education and social class; but the form or 
style of political expression — favoring force or persuasion, com- 
promise or arbitrary dictation, being tolerant or narrowly preju- 
diced, flexible in policy or rigidly dogmatic — is apparently largely 
determined by personality. At least this seems clear with regard to 
the political extremes. It is not yet certain whether the same char- 
acteristics make for extremism in all national groups and institu- 
tional settings, but that also seems highly likely. 

Prominent among the traits which make for extremism appear 
to be the following: exaggerated faith in powerful leaders and in- 
sistence on absolute obedience to them; hatred of outsiders and 
deviates; excessive projection of guilt and hostility; extreme cyni- 
cism; a sense of powerlessness and ineffectiveness (alienation and 
anqmie) ; suspicion and distrust of others; and dogmatism and 
rigidity. Some of these terms have been or will be shown to be 
merely alternative designations of the same phenomenon, but some 
such general syndrome of authoritarianism, dogmatism, and aliena- 
tion undoubtedly is the psychological root of that political extrem- 
ism which makes this type actively or potentially disruptive to 
democratic systems. 

If political extremism is indeed an accompaniment — and even 
more a product — of a certain personality syndrome, and if this syn- 
drome produces the equivalent extremism in all national popula- 
tions and subgroups, that fact poses a considerable challenge to 
the student of national character in its relation to political systems. 
At once we face this question: Are the societies which have a long 
history of democracy peopled by a majority of individuals who 


possess a personality conducive to democracy? Alternatively, are so- 
cieties which have experienced recurrent or prolonged authori- 
tarian, dictatorial, or totalitarian government inhabited by a 
proportionately large number of individuals with the personality 
traits we have seen to be associated with extremism? In other words, 
can we move from the individual and group level, to generalize 
about the relations of personality and political system at the societal 

Almost all the modern students of national character are con- 
vinced that the answer to this question is in the affirmative. Syste- 
matic empirical evidence for this faith is unfortunately lacking. 
To prove the point we would be required to show that the qualities 
of personality presumably supportive or less destructive of democ- 
racy are more widely prevalent in stable democracies such as the 
United States, England, Switzerland, or Sweden than in Germany, 
Japan, Italy, or Russia. At the present time we cannot offer such 
proof. We will continue to be unable to settle this question until 
we undertake nation-wide studies of modal personality patterns — 
such as we do of literacy or per capita income — and test their rela- 
tion to the forms of political organization in various countries. 
Before we undertake such studies we must have some conception of 
the character types for which we are looking. 

The problem of defining anything as broad as "the democratic 
character" may be much like the problem of locating the Manches- 
ter economists' "economic man" who Unamuno somewhere de- 
scribed as "a man neither of here nor there, neither this age nor an- 
other, who has neither sex nor country, who is, in brief, merely an 
idea — that is to say, a 'no-man.' " 

The danger of excessive generality in defining the democratic 
character is not greater than the danger of "misplaced concrete- 
ness," that is, defining the characterological requirements of atiy 
democracy as identical with those of some particular people who 
have a strong democratic tradition. For example, it has been true 
of the great majority of commentaries on the people of the United 
States, going back to its earliest days, that "practicality" and "em- 
phasis on religion" have been consistently cited as American traits 
(Coleman 1941 ) . Yet it would be difficult to argue that either qual- 
ity is a sufficient or even a necessary requirement for effective citi- 
zenship in a democracy. The same may be said of other traits fre- 
quently cited as characterizing the American people, such as 
valuing success and achievement, which are also strongly empha- 


sized in Japanese culture, or the marked emphasis on activity 
and work, which is also commonly cited as typifying the German 

While observing these cautions, we should not avoid postulating 
certain qualities which are probably indispensable to the long-run 
maintenance of a democratic political order. In holding this view 
we do no more than did De Tocqueville. De Tocqueville weighed 
the role of geography and climate, of religion and political institu- 
tions, and finally of what he called "manners," meaning thereby 
"various notions and opinions current among men . . . the mass of 
those ideas which constitute their character of mind . . . the whole 
moral and intellectual condition of a people." Comparing Mexico, 
South America, and the United States in these terms, he concluded: 
"The manners [character] of the Americans of the United States 
are the real cause which renders it the only one of the American na- 
tions that is able to support a democratic government ... I should 
say that the physical circumstances are less efficient than the laws, 
and the laws very subordinate to the manners [character] of the 
people" (1947:213). 

De Tocqueville's insistence that the maintenance of democracy 
depends upon the primacy of certain popular values, and what we 
would today call character traits, has often been reaffirmed since by 
numerous authorities including men as widely separated in formal 
philosophical allegiance as Sidney Hook and Jacques Maritain.'^ 
What specific qualities do we then require in a people as a neces- 
sary condition for the maintenance of a democratic political order? 
Even a casual content analysis of any sampling of opinion on the 
democratic society reveals an extraordinary degree of agreement 
about the values, attitudes, opinion and traits of character which are 
important to its maintenance. The various formulations may be 
summed up by reference to conceptions about others, about the 
self, about authority, and about community and society. 

Values about the Self. All authorities are agreed that demo- 
cratic societies require widespread belief in what Maritain calls the 
"inalienable rights of the person," and Hook "the belief that every 

' Hook has said, for example, "Democracy is an affirmation of certain attitudes and values 
which are more important than any particular set of institutions" (i9jo:294). Maritain argues 
that "the democratic impulse burst forth in history as a temporal manifestation of the gospel" 
and says directly that the democratic ideal "is the secular name for the ideal of Christianity" 
(1944:65). It does not seem necessary or desirable to clutter the text in the remainder of this 
section with source and page citations for each of the numerous quotations. In addition to the 
cited works of Hook and Maritain the main sources are Lasswell (1951) and De Tocqueville 


individual should be regarded as possessing intrinsic worth or dig- 
nity." "Where low estimates of the self are permitted to develop," 
says Harold Lasswell, "there the democratic character cannot de- 

Orientation toward Others. The basic dignity not only of the 
self but of all others is an essential ingredient cited by virtually 
every theory on the democratic character. This particularly mani- 
fests itself in the concept of equality, under which Hook includes 
recognition "that equal opportunities of development should be 
provided for the realization of individual talents and capacities." 
To hold this view one must have a basic acceptance of other people. 
In Lasswell's words: "The democratic attitude toward other hu- 
man beings is warm rather than frigid, inclusive and expanding 
rather than exclusive and constricting ... an underlying personal- 
ity structure which is capable of 'friendship' as Aristotle put it, 
and which is unalienated from humanity." Underlying these atti- 
tudes is a fundamental conception of the perfectibility of man, 
which De Tocqueville phrased as the belief "that a man will be 
led to do what is just and good by following his own interest rightly 

Orientation toward Authority. At the core of the democratic 
personality lies a stress on personal autonomy and a certain distance 
from, if not distrust of, powerful authority, or, to put it negatively, 
an absence of the need to dominate or submit such as is found in 
the authoritarian personality. As Sidney Hook phrased it: "a posi- 
tive requirement of a working democracy is an intelligent distrust 
of its leadership, a skepticism stubborn but not blind, of all demands 
for the enlargement of power, and an emphasis upon critical 
method in every phase of social life . . . Where skepticism is replaced 
by uncritical enthusiasm ... a fertile soil for dictatorship has been 
prepared." Almost identical language is used by Maritain. Maritain 
described the democratic philosophy as one insisting on the "po- 
litical rights of the people whose consent is implied by any political 
regime, and whose rulers rule as vicars of the people ... it denies to 
the rulers the right to consider themselves and be considered a su- 
perior race and wills nevertheless that their authority be respected 
on a juridical basis. It does not admit that the state is a transcendent 
power incorporating within itself all authority and imposed from 
above upon human life . . ." The same idea is stressed by Lasswell 
who says: "the democratic character is multi-valued rather than 
single valued . . . disposed to share rather than to monopolize. In 


particular, little significance is attached to the exercise of power 
as a scope value . . . [for] when the demand for respect is the con- 
suming passion, other values are sacrificed for the sake of receiving 
symbolic acknowledgments of eminence." 

Attitudes toward the Community. Although overweaning au- 
thority may be controlled, there is always the danger of that 
tyranny of the majority which De Tocqueville early warned might 
undo democracy. This realization has repeatedly led those who 
sought to define the democratic character to stress the importance 
of openness, ready acceptance of differences, and wilhngness to 
compromise and change. De Tocqueville early anticipated this 
point, as he did so many others. Stressing the belief "that every man 
is born of the right of self-government, and that no one has the 
right of constraining his fellow creatures to be happy," he went on 
to say we must recognize "society as a body in a state of improve- 
ment, [and] humanity as a changing scene in which nothing is or 
ought to be permanent." Hook also speaks of the importance of 
"a belief in the value of differences, variety, and uniqueness in a 
democracy [where] differences of interest and achievement must 
not be merely suffered, they must be encouraged." According to 
Hook this requires that the ultimate commitment of a democracy 
must be in some method by which value conflicts are to be resolved, 
which in turn means that policies must be treated as hypotheses, 
not dogmas, and customary practices as generalizations rather than 
as God-given truths. 

It will be apparent from this extremely brief review that there 
is substantial agreement about the core personal beliefs and values 
which have been frequently identified as important to the main- 
tenance of a democratic order. The relevant "themes" can, of 
course, be integrated into the personality at different levels. They 
may reflect opinions publicly held, but not vitally important to the 
person. They may represent basic attitudes or central values in the 
belief system, typical "ideologies" to which the individual has deep 
allegiance. Or they may be even more "deeply" embedded in the 
personality at the level of character traits and modes of psycho- 
dynamic functioning. Most of the outstanding writers on the demo- 
cratic character do not trouble to distinguish these "levels." I have 
not attempted above to sort them out, and merely note here that 
most of the characterizations given above are statements at the level 
of ideology. We can, however, translate or transform the classic 
portrait of the democratic character to present it in the language of 


clinical psychology, expressed in terms of character traits, defenses, 
ways of dealing with wishes and feelings, and the like. In those 
terms, the democratic character emerges at the opposite pole from 
the authoritarian personality syndrome. The citizen of a democracy 
should be accepting of others rather than alienated and harshly 
rejecting; open to new experience, to ideas and impulses rather 
than excessively timid, fearful, or extremely conventional with re- 
gard to new ideas and ways of acting; able to be responsible with 
constituted authority even though always watchful, rather than 
blindly submissive to or hostily rejecting of all authority; tolerant 
of differences and of ambiguity, rather than rigid and inflexible; 
able to recognize, control, and channel his emotions, rather than 
immaturely projecting hostility and other impulses on to others. 
This model of the democratic personality represents only a very 
rough first approximation. Although it is based on a great deal of 
philosophical wisdom and historical experience, by the standards 
of modern social science it rests on an extremely narrow and un- 
certain base of empirical research. Indeed, it might be argued that 
at the present moment there is no relevant evidence which meets 
the standards set by contemporary social science research. It is 
largely to the future that we must look for refinement of the model, 
and for testing of its actual relevance for political systems and 
popular participation in them. No doubt some elements in the 
model will be discarded, others added. It may even be discovered 
that some one element is critical, all the others incidental or even 
irrelevant. In the present stage of our work it is important to avoid 
premature closure through the exclusive concentration on one con- 
ceptual scheme for analyzing personality. It is true that earlier 
efforts which accepted publicly offered opinions, attitudes, and 
values as guides to the individual's probable political action were 
often naive and misleading. Nevertheless, an analysis couched ex- 
clusively in terms of psychodynamic depth psychology, of defenses, 
projective tendencies, and the like may also leave out much which 
is of great significance in shaping the pattern of political Hfe. We 
cannot be satisfied with a scheme of personality analysis which is 
insensitive to themes such as self-centeredness or "privatism" which 
Gillespie and Allport (1955) found so important in distinguishing 
the students from different countries in their study. Nor can we 
be content with an analysis of the "compulsive" German character 
(Kecskemeti 1947) if it leads us to neglect the feelings of obligation 
to self and society (McClelland 1958). 


Whatever the defects of the available scheme, the use of some 
explicit model is essential to focus our studies in this area. It is also 
a necessary condition for the meaningful comparison of different 
studies, and particularly for our efforts to cumulate the results in 
ever firmer generalizations or conclusions. We must particularly 
regret, therefore, that so few of the empirical investigations into the 
relations of character and political systems have sought systemati- 
cally to test the model of the democratic character presented above, 
or, for that matter, any other explicit model. 


With very few exceptions, the available studies of modal or group 
personality unfortunately suffer from several defects which make 
them poor evidence in support of any systematic proposition. As a 
rule they are not designed to test any theory or validate any model. 
They are usually based on very small and haphazardly selected sam- 
ples, making it extremely difficult to generalize with any confidence 
beyond the sample itself or the narrow circle from which it is drawn. 
In addition, the analysis is usually based on the total sample, with- 
out basic differentiation of the characteristics of subgroups, 
whether deviant or merely variant. More serious for our purposes 
is the fact that the description of personality is generally cast in 
clinical or psychodynamic terms which are difficult to relate to so- 
cial structure. Even in the rare cases when a study has given atten- 
tion to the more politically relevant realms of personality such as 
attitude toward authority, tolerance of ambiguity, acceptance of 
differences, and the need for power, it generally fails to record in- 
formation on the political attitudes and opinions, the party affili- 
ation, or other political characteristics of the subjects. Most of these 
studies, therefore, are obviously of limited usefulness to the student 
of politics. Only in the last few years have we attained the first, 
limited personality inventory of a representative sample of the na- 
tional population of the United States — and this applies only to the 
F scale, as we have already noted, and more recently to the TAT 
variables of n affiliation, achievement, and power.^ There are ap- 
parently no comparable results on these or any other dimensions for 
any other modern nation, and it will undoubtedly be many years 

^ The test was administered in connection with the national survey sponsored by the Joint 
Commission on Mental Illness and Health and conducted by the Survey Research Center of the 
University of Michigan. Reports on this material are in preparation by Gerald Gurin, Joseph 
Veroff, and John Atkinson. 


before we have such results for a number of major nations simul- 

Even when we attain good data on the distribution of personality 
traits in a number of national populations, a great many questions 
will remain. For example, we will need to understand better the 
relation between personality dispositions in the rank and file of a 
population, and their orientation to different kinds of leadership. 
The decisive factor affecting the chances of preserving democracy 
may not be the prevalence of one or another undemocratic per- 
sonality type, but rather the relation between the typical or aver- 
age personality and that of the leaders. It is highly unlikely that any 
character type will be found to be invariably associated with a single 
form of political system. Nevertheless, certain personality types 
may indeed be more responsive to one than to another form of gov- 
ernment. Their character, then, may be an important determinant 
of their susceptibility to certain kinds of influence. Thus, Dicks 
does not argue for the propensity toward authoritarian government 
per se in the German character. The typical German character de- 
lineated by Dicks was a type highly susceptible to the style of leader- 
ship the Hitler movement offered and extremely vulnerable to the 
kind of propaganda appeals it utilized. Much the same conclusion 
is suggested by Erikson's (1950) analysis of the German character 
and Hitler's appeal to it. Neither analysis should be interpreted as 
suggesting that the German character, as described, could not under 
any circumstances adjust to or function in any democratic politi- 
cal order. McClelland's analysis (1958) of the distinctive structure 
of obligations to self and society in Germany and the United States 
is particularly interesting for the light it throws on this question. 

Whatever the distribution of personality types, including leaders, 
in any population, we will want to know what produces the types. 
This enormously complex problem is one I have been obliged by 
limits of space to ignore almost entirely, although it is one of the 
most fundamental facing the field. The predominant opinion 
among students of national character is that these types arise mainly 
out of the socialization process, and that in democratic societies 
the family structure is one which generates individuals adapted to 
life in a democracy. The typical argument was forcefully stated 
by Ralph Linton when he declared: "Nations with authoritarian 
family structure inevitably seem to develop authoritarian govern- 
ments, no matter what the official government forms may be. Latin 
American countries with their excellent democratic constitutions 


and actual dictatorships would be a case in point" (1951:146). 

Linton's opinion is not uniformly held. On the basis of a thor- 
ough review of a great deal of relevant empirical research, Herbert 
Hyman (1959) poses a formidable challenge to this assumption 
and suggests a number of other factors — particularly experiences 
in adulthood — which may account for the political orientations we 
observe in certain groups. Even after we secure data on the distri- 
bution of personality characteristics in large populations, there will 
be much work to be done in discovering what produces the propen- 
sity to extremism, how it operates, and what — if anything — 
changes or modifies it. 

Another problem we must face is the relation between personal- 
ity factors and other forces which affect the political process (cf. 
Levinson 1958). To analyze political participation and political 
structures through a study of personality and its statistical distri- 
bution is, of course, only one of the possible avenues of approach to 
the problem. Clearly, political institutions and political action can 
not be comprehended exclusively or even predominantly by refer- 
ence to attitudes and values. The history of a people obviously plays 
a major role in shaping the basic structure of their political institu- 
tions. And institutional frameworks, once established, may have an 
endurance much greater than the formal allegiance to their prin- 
ciples would have indicated. Indeed, once firmly established, insti- 
tutions have the capacity to develop or generate support among 
those whose early disposition would hardly have led them to move 
spontaneously in that direction. 

A recent extensive comparative study by S. M. Lipset (1959) of 
the relation between a complex of factors including industrializa- 
tion, urbanization, literacy, education, and wealth, reveals that they 
are highly correlated not only with each other, but also with the 
existence of stable democratic systems.® None of these factors cited 
by Lipset is at all psychological or attitudinal, but it is interesting 
to note that in seeking to understand why these factors play such 
a role, Lipset had to fall back from these more ''objective" to more 
subjective causes, in particular to such concepts as the "effective- 
ness" and the "legitimacy" of a political system in the eyes of its 
constituents. By effectiveness he means the capacity to satisfy the 

*De Tocqueville made the same point: "Their ancestors gave [the people of the United States] 
the love of equality and of freedom, but God himself gave them the means of remaining equal 
and free by placing them on a boundless continent . . . When the people rules it must be 
rendered happy or it will overthrow the state, and misery is apt to stimulate it to those ex- 
cesses to which ambition rouses kings" (1947:185). 


basic interests of most members of society, or of the most important 
groups in it, and by legitimacy "the capacity of a pohtical system 
to engender and maintain the behef that existing pohtical institu- 
tions are the most appropriate or proper ones for the society" 
(1960:77). Surely the tolerance of ambiguity, the readiness for 
compromise, the level of projectivity characteristic of a people or 
important subgroups, will play a major role in shaping the "ef- 
fectiveness" of the political system and even its freedom of action 
to be effective. The value placed on autonomy versus control and 
direction, the strength of needs for power or achievement, the wish 
for dominance or subordination, the orientation toward authority 
figures, will all clearly play an important part in determining 
whether a particular political system is felt by people to be legiti- 
mate or not. 

Although further refinements are needed, it is not likely that we 
will make any further unusual leaps along the line of analysis which 
Lipset has so diligently pursued. By contrast, the role of psycho- 
logical factors — of attitudes, values, and character traits — in in- 
fluencing the political process is an almost virgin field which prom- 
ises a rich harvest. To secure it we must overcome imposing but by 
no means insuperable obstacles. We need to clarify our concepts, 
isolating or delineating those personal characteristics which, on 
theoretical grounds, seem to have the greatest relevance for the 
development and functioning of the political system. We must also 
refine our analysis of the political system, so that our descriptive 
categories are maximally analytical and conducive to comparative 
study. Our next step must be to assess systematically the distribu- 
tion of these qualities in different national populations and in im- 
portant subgroups of those populations. This poses one of the moist 
difficult methodological problems, since the meaning of important 
terms, the pattern of response to tests, and the interpretation of 
those responses are highly variable as we move from country to 
country. On this base we can then proceed to correlational and 
causal analyses of the relations between opinions, values, and per- 
sonality on the one hand, and the quality of political participation 
and the stability of political structures on the other. We may thus 
develop a comparative social psychology of the political process to 
support and supplement our traditional study of politics. 



Abel, Theodora M. and Francis L. K. Hsu 

1949 Some aspects of personality of Chinese as revealed by the Rorschach 
test. Rorschach Research Exchange and Journal of Projective Tech- 
niques 13:285—301. 

Adorno, T. W., E. Frenkel-Brunswik, D. J. Levinson and R. N. Sanford 

1950 The authoritarian personality. New York, Harper and Bros. 
Almond, Gabriel A. 

1950 The American people and foreign policy. New York, Harcourt Brace. 
1956 Comparative political systems. The Journal of Politics 18:391-409, 
AuBERT, ViLLEM, B. R. FiSHER and Stein Rokkan 

1954 A comparative study of teachers' attitudes to international problems 
and policies. Journal of Social Issues 10:25—39. 
Bauer, Raymond A. 

1953 Psychology of the Soviet middle elite. In Personality in nature, society, 
and culture, Kluckhohn, Murray, and Schneider, eds. New York, Alfred 
Benedict, Ruth 

1934 Patterns of culture. Boston, Houghton Mifflin. 

1943 Race: science and politics (rev. ed. ). New York, Viking Press. 
1946 The chrysanthemum and the sword. Boston, Houghton Mifflin. 
Bleuler, M. and R. 

1935 Rorschach's ink-blot test and racial psychology: mental peculiarities 
of Moroccans. Character and Personality 4:97—114. 

Boas, Franz 

1 9 10 Psychological problems in anthropology. American Journal of Psy- 
chiatry 21:371—384. 
Brickner, Richard M. 

1943 Is Germany incurable? Philadelphia, J. B. Lippincott. 
Brogan, D. W. 

1933 Government of the people, a study in the American political system. 
New York, Harper and Bros. 

1944 The American character. New York, Alfred Knopf. 
Buchanan, W. and H. Cantril 

1953 How nations see each other, a study in public opinion. Urbana, 111., 
University of Illinois Press. 

Campbell, Angus, G. Gurin and "W. E. Miller 

1954 The voter decides. Evanston, 111., Row, Peterson. 
Carstairs, G. M., R. W. Payne and S. Whitaker 

i960 Rorschach responses of Hindus and Bhils. Journal of Social Psychology 

51 :2i7— 227. 
Christie, Richard and M. Jahoda, ed. 

1954 The authoritarian personality: studies in continuities in social research. 

Glencoe, 111., Free Press. 


Coleman, Lee 

1 94 1 What is American: a study of alleged American traits. Social Forces 

CoMMAGER, Henry Steele, ed. 

1947 America in perspective, the United States through foreign eyes. New 
York, Random House, 

Da vies, James C. 

1954 Charisma in the 1952 campaign. American Political Science Review 
48:1083—1 102. 
Dicks, Henry V. 

1950 Personality traits and national socialist ideology, a war-time study of 

German prisoners of war. Human Relations 3:11 1— 1 54. 
1952 Observations on contemporary Russian behavior. Human Relations 
Eysenck, H. J. 

1954 The psychology of politics. London, Routledge and Kegan Paul. 

Fortes, Meyer and E. E. Evans-Pritchard, ed. 

1940 African political systems. London and New York, Oxford University 

Fromm, Erich 

1 94 1 Escape from freedom. New York, Farrar and Rinehart. 

Gillespie, James M. and Gordon W. Allport 

1955 Youth's outlook on the future: a cross-national study. Garden City, 
New York, Doubleday. 

GoRER, Geoffrey 

1943 Themes in Japanese culture. Transactions of the New York Academy 
of Sciences Ser. II, 5, 106-124. 

1948 The American people, a study in national character. New York, W. W. 

1955 Exploring English character. London, Cresset Press. 

GoRER, Geoffrey and John Rickman 

1949 The people of Great Russia, a psychological study. London, Cresset 

Haring, Douglas G. 

1948 Personal character and cultural milieu. Syracuse, N.Y., Syracuse Uni- 
versity Press. (3ded. 1956.) 

Hearnshaw, F. J. C. 

1940 Germany, the aggressor throughout the ages. London, W. and R. 

Herz, Frederick 

1944 Nationality in history and politics. London, Routledge and Kegan Paul. 

Honigmann, John J. 

1954 Culture and personality. New York, Harper and Bros. 

Hook, Sidney 

1950 Reason, social myths, and democracy. New York, Humanities Press. 



1936 Studien uber authoritiit und familie. Paris, Alcan. 
Hsu, Francis L. K. 

1953 Americans and Chinese: two ways of life. New York, Schuman. 
Inkeles, Alex and D. J. Levinson 

1954 National character: the study of modal personality and sociocultural 
systems. In Handbook of social psychology, vol. II, G. Lindzey, ed. 
Cambridge, Addison-Wesley. 

Inkeles, Alex, Eugenia Hanfmann and Helen Beier 

1958 Modal personality and adjustment to the Soviet socio-economic system. 
Human Relations 11:1—22. 

Jaensch, Erich R. 

1938 Der gegentypus. Leipzig, Barth. 

Janowitz, Morris and D. Marvick 

1953 Authoritarianism and political behavior. Public Opinion Quarterly 

Kardiner, Abram 

1945 The psychological frontiers of society. New York, Columbia University 

Kardiner, Abram and L. Ovesey 

195 I The mark of oppression: a psychosocial study of the American Negro. 
New York, Norton. 

Kecskemeti, Paul and Nathan Leites 

1947 Some psychological hypotheses on Nazi Germany. Journal of Social 
Psychology, I 1947 26:141—183; II 1948 27:91-117; III 1948 27: 
241—270; IV 1948 28:141 — 164. 

Klineberg, Otto 

1944 A science of national character. Journal of Social Problems 19:147-162. 
195 I Psychological aspects of international relations. In Personality and po- 
litical crisis, Alfred H. Stanton and Stewart E. Perry, eds. Glencoe, 
111., Free Press. 

Kluckhohn, Clyde and Henry Murray 

1953 Personality in nature, society, and culture. New York, Alfred Knopf. 
(Rev. ed., with David Schneider, 1956.) 

Kornhauser, Arthur, H. L. Sheppard and A. J. Mayer 

1956 When labor votes, a study of auto workers. New York, University 

Kornhauser, William 

1959 The politics of mass society. Glencoe, 111., Free Press. 

Krout, Maurice H. and Ross Stagner 

1939 Personality development in radicals: a comparative study. Sociometry 

Lane, R. E. 

1955 Political personality and electoral choice. American Political Science 
Review 49:173—190. 


Lasswell, Harold D. 

1930 Psychopathology and politics. In The political writings of Harold D. 

Lasswell. Glencoe, 111., Free Press, 195 1. 
195 1 Democratic character. In The political writings of Harold D. Lasswell. 

Glencoe, 111., Free Press. 
1959 Political constitution and character. Psychoanalysis and the Psycho- 
analytic Review 46:3—18. 
Leites, Nathan C. 

1948 Psychocultural hypotheses about political acts. World Politics i: 

102— 1 19. 
195 1 The operational code of the politburo. New York, McGraw-Hill. 
Lerner, Daniel 

1958 The passing of traditional society; modernizing the Middle East. Glen- 
coe, 111., Free Press. 
Levinson, Daniel J. 

1957 Authoritarian personality and foreign policy. Conflict Resolution i: 


1958 The relevance of personality for political participation. Public Opinion 
Quarterly 22:3—10. 

Levy, David M. 

195 1 Anti-Nazis: criteria of differentiation. In Personality and political crisis, 
Alfred H. Stanton and Stewart E. Perry, eds. Glencoe, 111., Free Press. 
Linton, Ralph 

1945 The cultural background of personality. New York, D. Appleton- 

195 1 The concept of national character. In Personality and political crisis, 
Alfred H. Stanton and Stewart E. Perry, eds. Glencoe, Illinois, Free 
LiPSET, S. M. 

1959 Some social requisites of democracy: economic development and po- 
litical legitimacy. American Political Science Review 53:69—105. 

i960 Pohtical man: the social bases of politics. Garden City, New York, 
LuRiE, "Walter A. 

1937 A study of Spranger's value-types by the method of factor analysis. 
Journal of Abnormal and Social Psychology 8:17-37. 
Madariaga, Salvador de 

1929 Englishmen, Frenchmen, Spaniards: an essay in comparative psychology. 
London, Oxford University Press. 

Mannheim, Karl 

1950 Freedom, power and democratic planning. New York, Oxford Uni- 
versity Press. 
Maritain, Jacques 

1944 Christianity and democracy. New York, Scribners. 
McClelland, David, J. F. Sturr, R. H. Knapp and H. W. Wendt 

1958 Obligations to self and society in the United States and Germany. 
Journal of Abnormal and Social Psychology 56:245-255. 


McCloskey, Herbert 

1953 Conservatism and personality. American Political Science Review 52: 

McGranahan, Donald V. 

1946 A comparison of social attitudes among American and German youth. 
Journal of Abnormal and Social Psychology 41:245-257. 

McGranahan, Donald V. and I. Wayne 

1948 German and American traits reflected in popular drama. Human Re- 
lations 1:429-455. 

Mead, Margaret 

1942 And keep your powder dry. New York, William Morrow. 
195 1 Soviet attitudes toward authority. New York, McGraw-Hill. 

Merriam, Charles E. 

1925 New aspects of politics. Chicago, University of Chicago Press. Selec- 
tions reprinted in H. Eulau, S. J. Eldersveld and M. Janowitz, PoUtical 
behavior. Glencoe, 111., Free Press, 1956. 

Michels, Robert 

1949 Political parties. Translated by Eden and A. Paul. Glencoe, 111., Free 

Mills, Charles W. 

1956 The power eHte. New York, Oxford University Press. 

Moore, Henry T. 

1925 Innate factors in radicalism and conservatism. Journal of Abnormal 
and Social Psychology 20:234-244. 

Morris, Charles W. 

1942 Paths of life; preface to a world religion. New York, Harper and Bros. 

Morris, Charles W. 

1956 Varieties of human value. Chicago, University of Chicago Press. 

Ortega y Gasset, Jose 

1937 Invertebrate Spain. New York, W. W. Norton. 

Peak, Helen 

1945 Observations on the characteristics and distribution of German Nazis. 
Psychological Monographs vol. 59, no. 6, whole no. 276. 

Ringer, Benjamin B. and David L. Sills 

1952-53 Political extremists in Iran. Public Opinion Quarterly 16:689-701. 

RoKEACH, Milton 

1956 Political and religious dogmatism: an alternative to the authoritarian 
personality. Psychological Monographs vol. 70, no. 18, whole no. 425. 


1949 Father land: a study of authoritarianism in the German family. New 
York, Columbia University Press. 

Sforza, Carlo 

1942 The real Italians: a study in European psychology. New York, Colum- 
bia University Press. 


Siegfried, Andre 

1927 America comes of age. New York, Harcourt Brace. 

1930 France: a study in nationality. New Haven, Yale University Press. 
195 I Approaches to an understanding of modern France. In Modern France, 
Edward M. Earle, ed. Princeton, Princeton University Press. 
Spranger, Eduard 

1928 Types of men. Tubingen, Max Neimeyer, Verlag-halle. 
Stoetzel, Jean 

1955 Without the chrysanthemum and the sword. New York, Columbia 
University Press UNESCO. 
Stoodley, Bartlett H. 

1957 Normative attitudes of Filipino youth compared with German and 
American youth. American Sociological Review 22:553—561. 
Stouffer, Samuel A. 

1955 Communism, conformity and civil liberties; a cross section of the 
nation speaks its mind. Garden City, N.Y., Doubleday. 


1947 Democracy in America. New York and London, Oxford University 
Wallas, Graham 

1908 Human nature in politics. London. Selections reprinted in H. Eulau, 
S. J. Eldersveld, and M. Janowitz, Political behavior. Glencoe, 111., Free 
Press, 1956. 
Wolfenstein, M. and N. Leites 

1950 Movies: a psychological study. Glencoe, 111., Free Press. 

chapter 7 



Northwestern University 

In approaching the subject of American national character, stu- 
dents have experienced some unusual difficulties. What they have 
done so far is either to present pictures of contradictions with little 
or no attempt to reconcile the opposing elements, or to construct 
models of what, in their view, ought to be, with little or no attempt 
to deal with what actually occurs. In this chapter I shall try to 
show that the difficulties are not insurmountable, that the contra- 
dictions, though numerous, are more apparent than real, and that, 
even the models of what ought to be, though different from reality, 
can be meaningful once we achieve a proper perspective. 

A Picture of Contradictions 

After comprehensive sampling of the literature from early times 
down to 1940, Lee Coleman lists the following as "American traits": 
"associational activity, democracy and belief and faith in it, belief 
in the equality of all as a fact and as a right, freedom of the indi- 
vidual in ideal and in fact, disregard of law — direct action, local 
government, practicality, prosperity and general material well- 
being, Puritanism, emphasis on religion and its great influence in 
national life, uniformity and conformity (Coleman 194 1 1498) . 

It is clear at once that this list of traits not only fails to give cog- 
nizance to such obvious facts as racial and religious prejudice, but 
the different traits mutually contradict each other at several points. 

* This chapter is based on a paper presented at the American Psychological Convention, 1959, 
Cincinnati, Ohio, as part of a symposium under the chairmanship of Dr. Fred J. Goldstein 
of Los Angeles Psychiatric Service. I am greatly indebted to Donald T. Campbell, Millard Hoyt, 
Thomas Gladwin, and Melford Spire for their valuable criticism of this chapter. 



For example, values attached to "local government" and "democ- 
racy" are in direct contradiction to that of "disregard of law" 
leading to "direct action." The beliefs in "equality" and in "free- 
dom" are in direct contradiction to the emphasis on "uniformity 
and conformity." 

Cuber and Harper, writing nearly ten years later in a book en- 
titled Problems of American Society: Values in Conflict, have re- 
duced the total number of American values enumerated but not 
done much else. Their list is as follows: "monogamous marriage, 
freedom, acquisitiveness, democracy, education, monotheistic re- 
ligion, freedom and science" (Cuber and Harper 1948 : 369) . Cuber 
and Harper recognize that some of these values are inconsistent 
with each other and with social reality. But they attempt to ex- 
plain such inconsistencies as follows: 

On the surface it might seem relatively easy for a society, and especially for 
some one person, to discover such inconsistencies as these, evaluate the two 
positions, choose one, and discard the other .... But in practice it seems not 
to be so easy an undertaking. In the first place, logical inconsistency may con- 
stitute social consistency — that is, a person whose values seem inconsistent when 
analysed by a third party may regard himself to be quite consistent. Both values 
seem to him to be quite tenable because he can point out the other persons in 
the society as authority for the Tightness of each position. (Cuber and Harper 

As we shall see later, their explanation contains the germ of truth 
as to why the individual is not free to act as he sees fit, to make his 
value orientation more self-consistent, but it has not gone far 
enough. If every individual adheres to his inconsistent values be- 
cause he can resort to "other persons in the society as authority for 
the rightness of each position," then we cannot possibly explain 
how values in America would ever undergo change, and how some 
individuals are more affected by the inconsistencies than others, 
enough for them to espouse certain "causes" and throw their weight 
behind crusades for emancipation of the slaves or to bust up saloons. 

Over the years the analysis of American values has remained stag- 
nant at this level. Thus, in American Society Robin Williams again 
gives us no more than a catalogue of American values as follows: 
"achievement" and "success," "activity" and "work," "moral orien- 
tation," "humanitarian mores," efficiency and practicability, 
"progress," material comfort, equality, freedom, external conform- 
ity, science and secular rationality, nationalism-patriotism, de- 
mocracy, individual personality, racism and related group-superi- 


ority themes. (The quotation marks appHed to seven of these values 
areWilhams') (Wilhams 195 1 : 388-440; 1960:415-470). 

Wilhams does realize, perhaps more than the other authors, that 
the values are not of equal importance and that they have to be 
somehow related and reconciled with each other. Accordingly, in 
his conclusion on value orientation, he makes a summary classifi- 
cation to emphasize some and to de-emphasize others: 

a) Quasi values or gratifications: such as material comforts. 

b) Instrumental interests or means values: such as wealth, power, work, and 

c) Formal universalis tic values of western tradition: rationaUsm, impersonal 
justice; universaUstic ethics, achievement, democracy, equality, freedom, 
certain religious values, and values of individual personality. 

d) Particularistic, segmental or localistic values: best exemplified in racist- 
ethnic superiority doctrines and in certain aspects of nationalism (Wil- 
liams 1951:441; 1960:468-469). 

This classification accomplishes little. It is not simply a question 
of differences between professed values and the actual reality. Such 
differences are likely to be found in any society. More specifically 
the question is one of unresolved and unaccounted for differences 
between certain professed values and other professed values. We 
may reconcile "efficiency" as a value with the continuous blocking 
of modern improvements in the building trades as a matter of dif- 
ference between theory and practice. But how do we reconcile the 
"value of individual personality" with the oppressive and increasing 
demand for "conformity"? The most glaring contradiction exists 
between "equality," "freedom," and so forth on the one hand and 
"racist-ethnic superiority doctrines and certain aspects of national- 
ism" on the other. Williams tries to expunge the "ethnic superiority 
doctrines and so forth" by inaccurately classifying the latter as 
"particularistic, segmented or localistic values." 

It is easy to see how Williams errs here. If the belief in racist- 
ethnic superiority were truly segmental or localistic (by which I 
think Williams means that it is particular to the South) , how can 
we explain the racism that is also prevalent in the North? In fact, 
it has been aptly observed, and I think with some justification, that 
the only difference between the South and the North in the matter 
of racial attitudes is that the South is more open and honest about 
it, while the North is more covert and hypocritical about it. Of 
course, this view fails to consider the fact that the law by and large 
still supports racism in some Southern states, while the law is against 


it in the North. Besides, practically all the broad legislative and 
judiciary improvements affecting race relations have originated 
from the North. These legal changes do not, however, erase the 
widespread social, economic, and other forms of discrimination 
which are practiced in the North as well as in the South. Further- 
more, even if we say that the racist attitude is only characteristic 
of the South, we must inevitably be confronted with the question: 
How does the South reconcile its racist attitudes with its professed 
belief in democracy? Are the North and the South two funda- 
mentally separate cultures? 

Some students frankly take the line of least resistence by char- 
acterizing the American culture as "Schizoid" (Read Bain 1935: 
266-y6) , or inherently "dualistic," that is to say, full of opposites 
(Harold J. Laski 1948:738). This is the same sort of conclusion 
reached by Gunnar Myrdal who, after a mammoth investigation of 
the Negro-White relations, left the entire matter as An American 
Dilemma (1944). Apart from presenting many factual details on 
racial discrimination in this society, Myrdal said nothing more than 
that there is the problem of a psychological conflict between the 
democratic ideal of equality, on the one hand, and the existing in- 
equalities in race relations, education, income distribution, health 
benefits, and so forth, on the other. The few anthropologists who 
have bothered to study American values have hardly improved on 
this state of affairs. Thus, Kluckhohn expressed himself in 1941 
on this subject: 

While the relative unanimity over some kind of aid to Britain demonstrates 
that at least in a crisis a nexus of common purposes is still effective, the diagnostic 
symptom of the sickness of our society is the lack of a unifying system of canons 
of choice, emotionally believed in as well as intellectually adhered to. (Kluck- 
hohn 1941:175) 

When Kluckhohn gave us his more intensive analysis of the 
American culture six years later, we can readily understand why his 
early conclusion on American values was as it was. Because his 
analysis consists of another list of "orientations" and "suborienta- 
tions" that are very much in the manner of Robin Williams' treat- 
ment detailed above on pages 211 and 212 (Kluckhohn and 
Kluckhohn 1947). 

Thus, our understanding of American values is today no better 
than it was several decades ago. Periodically we note the conflicts 
and inconsistencies among the different elements, but we leave them 
exactly where we started. 


An American Blind Spot 

I have taken so much time to come to this futile point because I 
do not wish to be accused of setting up a nonexistent straw man and 
then, with the flourish of discovery, knock him down. 

The reason for this lack of progress in the scientific analysis of 
value conflicts inherent in American culture is, I believe, to be 
found in the fact that many Western and especially American 
scholars have been too emotionally immersed in the absolute good- 
ness of their own form of society, ethic, thought, and religion that 
it is hard for them to question them, even in scientific analyses. 
Consequently, they cannot see anything but the eventual triumph 
of their cultural ideals such as freedom and equality over realities 
such as racism and religious intolerance. Some frankly see the 
former as the basic American values and the latter as outright devi- 
ations which need not even be considered. This attitude is most 
decidedly characteristic even of eminent scholars of American his- 
tory such as Henry Steele Commager. In his book The American 
Mind he practically dismisses the Negro and, in fact, all nonwhites 
with one sentence: 

Nothing in all history had ever succeeded like America, and every American 
knew it. Nowhere else on the globe had nature been at once so rich and so 
generous, and her riches were available to all who had the enterprise to take them 
and the good fortune to be white (1950:5). 

I would have regarded the last sentence quoted here to be Com- 
mager's satire on the prevailing attitude of the American public, if 
not for the fact that, in the rest of his 443 pages, he makes no more 
than a few passing references to the treatment of Negroes (in one 
of which the word "Oriental" is inserted) . Furthermore, in these 
references, the Negroes might well have been as important as the 
wayside flowers trampled on by the horses drawing westward 
wagons driven by white Americans. When Commager comes to 
twentieth century America, he seems to be most exasperated 
by the adverse manifestations of the American mind in the form 
of crime, racial and religious bigotry, lawlessness, irreligion, loose- 
ness of sex mores, conformity, class formation, and so forth. He 
seems so intent upon denying them, yet cannot, that he speaks in 
the following confusing vein : 

All this presented to the student of the American character a most perplexing 
problem. It was the business of the advertisers to know that character, and their 
resources enabled them to enlist in its study the aid of the most perspicacious 


sociologists and psychologists. Yet if their analysis was correct, the American 
people were decadent and depraved. No other evidence supported this conclusion. 
Advertisers appealed to fear, snobbery, and self-indulgence, yet no one familiar 
with the American character would maintain that these were indeed its pre- 
dominant motivations, and statesmen who knew the American people appealed 
to higher motives, and not in vain. The problem remained a fascinating one, 
for if it was clear that advertisers libeled the American character, it was equally 
clear that Americans tolerated and even rewarded those who libeled them. (Com- 
mager 1944:419; italics mine) 

Besides its obvious one-sidedness (for example, his statement that 
"the statesmen who knew the American people appealed to higher 
motives, and not in vain" is about as true as another which reads, 
"the statesmen who knew the American people appealed to baser 
motives, and not in vain,") , Commager contradicts himself badly. 
Unable to deny the reality of facts uncovered by scientists, facts 
which are used profitably by advertisers, yet unable to bring him- 
self to see them in their true perspective, he solved his academic 
dilemma by branding the facts as "libel." 

Gordon Allport commits the same error in his book The Nature 
of Prejudice. In its entire 519 pages Allport theorizes about mankind 
and religion, but his mankind is Western mankind (where he occa- 
sionally refers to Negroes and Orientals, he is merely speaking about 
to what different extents the different Western groups reject them) , 
and by religion he means Protestantism, Catholicism, and Judaism, 
with nothing even about Eastern Orthodoxy and one sentence on 
Islam. Limited by such a culture-bound framework Allport is not 
unnaturally inconsistent (1954). In discussing racial prejudice, 
Allport relies heavily on experimental psychology. There is a great 
deal of evidence that the more prejudiced personality tends to be 
one which is more in need of definiteness and more moralistic. For 
example, "he is uncomfortable with differentiated categories; he 
prefers them to be monopolistic" (Allport 1954:175, 398-408). 
Here Allport apparently accepts the conclusion to which his evi- 
dence leads him. However, in connection with religious bigotry 
Allport seems to adopt a different procedure altogether. Here he 
first admits that religions which claim to possess final truths are 
bound to lead to conflicts, and that individuals who have no re- 
ligious affiliations tend to show less prejudice than do church mem- 
bers. But these are, in his words, too "distressing" to him and so de- 
mands "closer inspection" (Allport 1954:451). 

To the student, what Allport means by "closer inspection" turns 
out to be a surprise, for Allport departs from the acceptable prin- 


ciple of science by purposely attempting to negate stronger evi- 
dences in favor of much flimsier facts. He admits that, quanti- 
tatively, the correlation between greater church affiliation and 
greater prejudice is correct, but he also insists that it is not correct 
because there are "many cases" where the influence of the church 
"is in the reverse dirction" (Allport 1954:451). In other words, 
Allport finds the evidences too distressing because they show the 
Christian churches and the Christian values in an unfavorable light. 
He simply cannot tolerate the fact that the absolutist Christian 
faith and the exclusive Christian church membership do lead to 
greater prejudice. Under the circumstances, Allport has no alterna- 
tive but to throw overboard the quantitative evidence in favor of 
some qualitative statements. 

Yet even so sophisticated a social scientist as Lloyd Warner is no 
exception. In his book American Life, Dream and Reality he finds 
the Jonesville grade school children's evaluation of one another to 
be so strongly reflective of social-class values as to blind them to the 
actual reality (for example, children from the top classes were 
rated 22 times cleaner than those from the bottom, but in fact the 
latter as a whole came to school cleaner and neater than the former) . 
However, he also finds that the Jonesville high school children, 
though following a similar pattern, do not make such categorical 
and rigid judgments by class values. Warner's explanations of this 
difference are most revealing: 

Since the older children are presumably more the products of their culture 
than the younger ones, there appears to be a contradiction here. . . . Actually, 
the reasons for the differences in judgment help verify our hypothesis. The 
children in the high school, being products of American society, have learned 
to be less open and more careful about what they say and how they feel on the 
tabooed subject of status. Furthermore, they have learned to use American values 
of individtialism and are able to make clearer discriminations about the worth 
of an individual than are the younger children. (Warner 1953:182-3; italics 

The interesting thing is that Warner's second explanation here 
not only contradicts the one preceding it but contradicts his entire 
thesis, which is that social class values strongly influence American 
behavior and ideas. It is as though this second explanation came out 
by accident, perhaps a Freudian slip of his research pen, for in senti- 
ments like "the worth of the individual" many Americans find 
real emotional security. 

What we have to see is that in the minds of a majority of our 


scholars the idea of democracy and Christianity, with their respec- 
tive attributes of freedom and equahty in one case and of love and 
mercy in the other, are the over-all American values par excellence. 
They are so consciously upheld that all explanations of American 
behavior must somehow begin and end with it. Any evidence con- 
trary to this mold is therefore treated as deviation or as "regional 
phenomena," as "libel," as creating a "schizoid" situation, a "di- 
lemma." This in my view is the blind spot to many of our Western 
social scientists today. Given this blind spot, our scientists have 
consistently confused what ought to be with what is. It leads many 
scholars to explain the kind of American behavior they deem de- 
sirable by one theory, and another kind of American behavior, 
which they abhor and which contradicts the first kind, by 
another and contradictory theory. Some even misuse the eclectic 
approach by pleading the multiplicity of correlates or causation in 
complex human affairs. 

The fundamental axiom of science is to explain more and more 
facts by fewer and fewer theories. Anyone can explain all charac- 
teristics of a given situation with as many different theories, but 
his explanation will not be of value as a piece of work of science. 
It might be close to a factual description. Or it might be close to 
fantasy or rationalization. The axiom of explaining more and more 
facts by fewer and fewer theories is especially crucial if the facts 
are obviously related, as when they occur in the same organized 
society and often among and in the same individuals. 

Once this is admitted it becomes obvious that, when confronted 
with contradictions in the object of his inquiry, the scientist's first 
duty is, instead of trying to treat them as discrete entities and ex- 
plaining them with contradictory hypotheses, to explore the possi- 
bility of a link between the contradictory phenomena. In doing so 
the scientist is not presuming that values in any given society must 
be totally consistent with each other and that all contradictions 
must be resolved. It is perfectly possible that many societies, being 
large and complex, have inconsistent or contradictory values. But 
what our scientists so far would seem to fail or even refuse to do is 
to concede even the possibility of any positive connection between 
these contradictory values. 

Self-Relionce, Fear of Dependency, and Insecurity 

What we need to see is that the contradictory American "values" 
noted by the sociologists, psychologists, and historians are but mani- 


festations of one core value. Furthermore, many scholars must have 
been aware of this core value in one way or another but, because of 
their blind spot, have failed to recognize its importance. The Amer- 
ican core value in question is self-reliance, the most persistent psy- 
chological expression of which is the fear of dependence. It can be 
shown that all of the "values" enumerated thus far, the mutually 
contradictory ones and the mutually supportive ones, the evil ones 
as well as the angelic ones, spring or are connected with self-reliance. 

American self-reliance is basically the same as English individual- 
ism except that the latter is the parent of the former while the 
former has gone farther than the latter. However, self-reliance 
possesses no basic characteristics which were not inherent in indi- 
vidualism. Individualism developed in Europe as a demand for po- 
litical equality. It insists that every individual has inalienable and 
God-given political rights which other men cannot take away and 
that every man has equal right to govern himself or choose his own 
governors. Self-reliance, on the other hand, has been inseparable 
in America from the individual's militant insistence on economic, 
social, and political equality. The result is while a qualified indi- 
vidualism, with a qualified equality, has prevailed in England and 
the rest of Europe, what has been considered the inalienable right 
of every American is an unlimited self-reliance and an unlimited 

It is not suggested here that all Americans do in fact possess the 
unlimited economic and social equality in which they firmly be- 
lieve. But it is easy to observe how strongly and widely the belief 
in them manifests itself. For example, the English have been able to 
initiate a sort of socialism in reality, as well as in name, but Ameri- 
cans, regardless of social security, farm subsidies, and other forms 
of government planning, intervention, and assistance, are as firmly 
as ever committed to the idea of free enterprise and deeply intol- 
erant toward other social systems. Similarly, the English still tend 
to respect class-based distinctions in wealth, status manners, and 
language, while Americans tend to ridicule aristocratic manners or 
Oxford speech, and resent status so much that Lloyd Warner, for 
example, describes it as being a "tabooed" subject in discussing 
Jonesville high school students. Finally, the English still consider 
the crown a symbol of all that is best and hereditary, Americans 
criticize the personal taste of their highest officials and at least 
have the common verbal expression that everybody can be presi- 


This self-reliance is also very different from self-sufficiency. Any 
Chinese or European village can achieve self-sufficiency as a matter 
of fact. The average self-sufficient Chinese farmer will have no feel- 
ing whatever about other people who are not self-sufficient. But 
American self-reliance is a militant ideal which parents inculcate in 
their children and by which they judge the worth of any and all 
mankind. This is the self-reliance about which Ralph Waldo Emer- 
son has written so eloquently and convincingly in some immortal 
pieces. This is also the self-reliance taught in today's American 
schools. The following is a direct quotation from a statement of 
"basic beliefs" given to the students by the social science depart- 
ment of one of the nation's best high schools in 1959: 

Self-reliance is, as it has always been, the key to individual freedom, and the 
only real security comes from the ability and the determination to work hard, 
to plan, and to save for the present and the future.^ 

American self-reliance is then not new. As a concept it is in fact 
well known and well understood. Yet such is the power of the blind 
spot that its over-all and basic importance has so far escaped our 
scientific attention. How the individualism of Western Europe has 
been transformed into American self-reliance is a question outside 
the scope of this paper. It has been dealt with elsewhere (Hsu 
1953:111-114). Suffice it to say here that under this ideal every 
individual is his own master, in control of his own destiny, and will 
advance and regress in society only according to his own efforts. He 
may have good or bad breaks but, 

Smile and the world smiles with you. 
Cry and you cry alone. 

It is, of course, obvious that not all Americans are self-reliant. 
No ideal of any society is uniformly manifested in all its members. 
But a brief comparison will make the point clearer. A man in tra- 
ditional China with no self-reliance as an ideal may not have been 
successful in his life. But suppose in his old age his sons are able to 
provide for him generously. Such a person not only will be happy 
and content about it, but is likely also to beat the drums before all 
and sundry to let the world know that he has good children who 
are supporting him in a style to which he has never been accus- 
tomed. On the other hand, an American parent who has not been 
successful in life may derive some benefit from the prosperity of 

^A mimeographed sheet issued to its pupils by a school in the Greater Chicago area, 1959. 


his children, but he certainly will not want anybody to know about 
it. In fact, he will resent any reference to it. At the first opportunity 
when it is possible for him to become independent of his children 
he will do so. 

Therefore, even though we may find many individuals in tra- 
ditional China and elsewhere who are in fact self-sufficient, and 
even though we may find individuals in America who are in fact 
dependent upon others, the important thing is to realize that where 
self-reliance is not an ideal, it is neither promoted nor a matter of 
pride, but where it is an ideal, it is both. In American society the 
fear of dependence is so great that an individual who is not self- 
reliant is an object of hostility and called a misfit. "Dependent 
character" is a highly derogatory term, and a person so described 
is thought to be in need of psychiatric help. 

However, it is obvious that no individual can be completely self- 
reliant. In fact, the very foundation of the human way of life is 
man's dependence upon his fellow men without which we shall have 
no law, no custom, no art, no science, and not even language. It is 
not meant that an individual human being cannot be trained, from 
the beginning of his life, to form no relationship with any fellow 
human being. But if an individual wishes to lead a human existence, 
in this society or any other, he is bound to be dependent upon his 
fellow human beings intellectually and technologically as well as 
socially and emotionally. Individuals may have differing degrees 
of needs for their fellow human beings, but no one can truly say 
that he needs no one. It seems that the basic American value orienta- 
tion of self-reliance, by its denial of the importance of other human 
beings in ones' life, creates contradictions and therefore serious 
problems, the most uniquitious of which is insecurity. 

This insecurity presents itself to the individual American in a 
variety of ways. Its most important ingredient is the lack of perma- 
nency both in ones' ascribed relationships (such as those of the fam- 
ily into which one is born) and in one's achieved relationships (such 
as marital relationship for a woman and business partnership for a 
man) . Its most vital demand on the individual is to motivate him in 
a perpetual attempt to compete with his fellow human beings, to 
belong to status-giving groups, and, as a means of achieving these 
ends, to submit to the tyranny of organization and to conform to 
the customs and fads of the peer group which are vital to his climb- 
ing and/or status position at any given time and place. In other 
words, in order to live up to their core value orientation of self- 


reliance, Americans as a whole have to do much of its opposite. 
Expressed in the jargon of science, there is, for example, a direct re- 
lationship between self-reliance and individual freedom on the one 
hand and submission to organization and conformity on the other 
(Hsu i960: 15 1 ) . Exactly the same force can be seen to link: 

a) Christian love with religious bigotry. 

b) Emphasis on science, progress, and humanitarianism with parochialism, 
group-superiority themes and racism. 

c) Puritan ethics with increasing laxity in sex mores. 

d) Democratic ideals of equality and freedom with totalitarian tendencies 
and witch hunting. 

These four pairs of contradictions are not exclusive of each other. 
For example, Christian love is in sharp contrast with racism as with 
religious bigotry. Similarly emphasis on science, and so forth, is as 
opposed to totalitarian tendencies and witch hunting as to paro- 
chialism and group superiority themes. In fact, we can contrast the 
first half of any of the above pairs with the second half of any other. 

Christian Love versus Christian Hate 

For the purpose of this paper we shall consider some of these 
contradictions in a composite whole: the American emphasis on 
Christian love, and freedom, equality, and democracy on the one 
hand, and racism and religious bigotry on the other. This is a con- 
tradiction which has tested the energy of some of the best euphemis- 
tic orators and the ingenuity of some of the most brilliant scholars. 
Especially in the religious area they try to write off the religious 
wars. They try to forget about the Holy Inquisitions. They try to 
ignore the hundreds of thousands of witches convicted and burned 
on the stake. They try to deny any connection between any of these 
and the Nazi Germany slaughter of the Jews, especially the anti- 
Semitism, anti-intellectualism, and racial persecution found, here 
covertly and there openly, in the United States. But when some 
scholars do realize that the past patterns are very much alive at pres- 
ent, though the specific techniques have changed, they tend to make 
harmless observations of which the following is a typical example: 

Worship in common — the sharing of the symbols of religion — has united hu- 
man groups in the closest ties known to man, yet religious differences have helped 
to account for some of the fiercest group antagonisms. (Elizabeth K. Notting- 
ham 1954:2) 

Williams, who quotes the above passage, goes a little further by 
suggesting two clues to the riddle as to why some worship in com- 


mon has united people and some has divided them : (a) "Not all con- 
flicts in the name of organized religion are actually "religious" and 
(b) there may be different degrees of involved commitment actu- 
ally at work in "nominal religious affiliations" (Robin M. Williams 
1956:14—15). But there is no observable basis for distinction be- 
tween "true" religious conflict and religious conflicts which are 
only nominally religious. Are theological controversies purely re- 
ligious or nominally religious? The truth is that, even if the conflict 
is over nothing but liturgy, or over the question of virgin birth, 
they are still fought between human beings each with personal, emo- 
tional involvements in specific issues. 

Williams' second clue is a more sound one. Put it differently, this 
is that the more "involved commitment" actually at work in nomi- 
nal religious affiliations the more religious dissension and bigotry 
there will be. Since the stronger one's commitment to an object or 
issue the more inflexible this commitment becomes, it is natural that 
more "involved commitment" will lead to more dissension and 
bigotry. Certain data quoted by Allport, referred to before, directly 
support this proposition.^ It is interesting to note that Williams, 
after stating this proposition, dismisses it as "extreme." Instead he 
collects a conglomeration of twenty divergencies in value — orien- 
tation which, he believes but does not demonstrate, are partially 
the basis of religious conflicts in the United States (Williams 1956: 

It is unnecessary to probe into the reasons why Williams attaches 
so little significance to his second clue. It is also beyond the scope of 
this paper to detail the irrelevancy of some of his "divergencies" to 
this problem on hand. We can, however, indicate how the link be- 

^ "Over four hundred students were asked the question, 'To what degree has religion been an 
influence in your upbringing?' Lumping together those who report that religion was a marked 
or moderate factor, we find the degree of prejudice far higher than among those who report that 
religion was a slight or non-existent factor in their training. Other studies reveal that indi- 
viduals having no religious afiSliation show on the average less prejudice than do church mem- 
bers." (Allport 1954:451) 

And again, "First, it is well to be clear concerning the existence of certain natural, and 
perhaps unresolvable, conflicts inherent in various aspects of religion. 

"Take first the claim of certain great religions — that each has absolute and final possession 
of Truth. People who adhere to different absolutes are not likely to find themselves in agreement. 
The conflict is most acute when missionaries are actively engaged in proselytizing divergent sets 
of absolutes. Moslem and Christian missionaries in Africa, for example, have long been at odds. 
Each insists that if its creed were completely realized in practice, it would eliminate all ethnic 
barriers between men. So it would. But in actuality, the absolutes of any one religion have never 
yet been accepted by more than a fraction of mankind. 

"Catholicism by its very nature must believe that Judaism and Protestantism are in error. 
And varieties of Judaism and Protestantism feel keenly that other varieties of their own faith 
are perverse in many points of belief." (Allport 1954:444—445) 


tween the degree of involved commitment in nominal religious af- 
filiations and the extent of dissension and bigotry is the source of the 
contradiction: Christian love versus Christian hate. It is not hard 
for the trained social scientist to note that religious affiliation in the 
United States today has become so largely a matter of associational 
affiliation that "the values that inhere in group affiliation and par- 
ticipation" far and above overshadow "the specific values espoused" 
by the religious body ( Williams 1956:17). The overwhelming proof 
of this is to be found in well-known works such as the Lynds' on 
"Middle Town" and Lloyd Warner and associates on "Yankee City" 
and "Jonesville," ^ but particularly in the results of a poll of 100,000 
Protestant ministers in all parts of the United States by the Chris- 
tian Century magazine in 195 1, to determine the "outstanding" and 
most "successful" churches. This poll showed twelve to be the 
chosen ones. One of the twelve was the First Presbyterian Church 
of Hollywood. 

The applauded "qualities" of this church have been analyzed else- 
where (Hsu 1953:273-277) . Suffice it to say here that the "success- 
ful qualities" of this church seem to be that the "happiness" of the 
parishioners revolves about the social and material endeavors which 
rebound to their benefit alone but that the spiritual faith and the 
quality of the ministers' teachings receive practically no attention. 

All this is understandable once we appreciate the persistent de- 
mands that the core American value of self-reliance makes on the 
individual. The churches must compete and, in order to exist and to 
be "successful," must satisfy the status quest of its members. To 
achieve that "success," the churches not only have to conform to 
the trend toward organization, but they must try to find new ways 
of increasing their memberships so as to reach greater "successes." 

In this psychology we can now find the common ground between 
religious bigotry and racial prejudice. Western religious dissensions 
have been associated with many things but their principal and per- 
ennial feature has been the search for original purity in ritual and 
belief. The Reformation was based on it. The entire evolution of 
Protestantism from the Lutheran church to Quakerism has had it 
as the central ingredient. The Holy Inquisition was instituted to 
ferret out impurity in Christian thought and practice. This fervent 

* Commenting on religion George C. Romans says: "We are apt to think that the choice 
of a church among people brought up in the Protestant tradition is a matter of individual 
conscience. No doubt it is. But it is certainly also true that the membership of churches, in 
Hilltown as in Boston, tended to correlate roughly with that of certain social groups" (1950:346). 


search for and jealous guard over purity expresses itself in the racial 
scene as the fear of genetic mixing of races which feeds the segre- 
gationist power in the North as well as in the South, no matter what 
rhetoric and other logic are employed. When religious affiliations 
have become largely social affiliations, this fear of impurity makes 
religious and racial prejudices undistinguishable. Religion is not 
the question. The point of the greatest importance is affiliation. The 
neighborhoods and clubs are as exclusive as the churches and church 
activities tend to be, in spite of all protestation of equality, democ- 
racy, worth of the individual. Christian love, and humility. 

The individual who is enjoined to be self-reliant, unlike one who 
is taught to respect authority and external barriers, has no perma- 
nent place in his society. Everything is subject to change without 
notice. He is always anxious to look above for possible openings to 
climb, but he is at the same time and constantly threatened from 
below by possible upward encroachment. In his continuous effort 
at status achieving and maintaining, the self-reliant man fears noth- 
ing more than contamination by fellow human beings who are 
deemed inferior to him. This contamination can come about in di- 
verse forms: sharing the same desks at the same schools, being 
dwellers of the same apartments, worshipping in the same churches, 
sitting in the same clubs, or being in any situation of free and equal 

In this context, as in others, individuals will vary in the extent 
to which they are pressed by the fear of inferiority. Some will join 
hate organizations, lynching mobs, and throw stones at Negro resi- 
dences or paint swastikas on Jewish synagogues. These are violent 
acts of prejudice. Others will do everything they legally or by de- 
vious means can do to keep individuals of certain religious, racial, or 
ethnic groups out of residential areas, certain occupations, and so- 
cial fraternities. These are active nonviolent acts of prejudices. Still 
others will quietly refuse to associate with members of religious, ra- 
cial, or ethnic minorities and teach their children to observe this 
taboo because one just does not do such things. These are passive non- 
violent acts of prejudice. 

Under such circumstances many, perhaps most, individuals find 
it impossible to act in the same way as they have professed and been 
taught. It is not that they love contradiction or that they are, ac- 
cording to their critics, hypocritical. It is simply that they are op- 
pressed by fears for losing satus — fears deeply rooted in a relatively 
free society with a core value of self-reliance. This is also why inte- 


gration of minorities, be they racial or religious, cannot reach a 
satisfactory destination either along the line of total assimilation 
into the majority way of life or along that of pluralism. There is 
some factual indication that Jewish youngsters who are raised as 
non- Jews have a much harder time to adjust to their peers in college 
than those who have been raised consciously and militantly to culti- 
vate their identity in Judaic tradition and church life. In other 
words, their complete identity and assimilation as Americans is al- 
ways subject to rejection (Samuel Teitelbaum 1953 ) .^ On the other 
hand, the rationalization in support of anti-Oriental legislation was 
that the Oriental standard of living was too low and that they were 
incapable of assimilation to the American way of life. 

A reverse proof of the hypothesis advanced in this paper is not 
hard to find. We have only to look at societies where obedience to 
authority and dependence relationship are encouraged and where 
the individual is not subject to such pressures coming with self- 
reliance and, therefore, more sure of his place in society. Individuals 
in such societies tend to have much less need for competition, status 
seeking, conformity, and, hence, racial and religious prejudices. For 
example, religious dissentions, persecutions, and conflicts have al- 
ways been prominent in the West as they have alway been rare in 
the Orient. In Japan and China, the few occasions on which reli- 
gious persecutions took place were invariably of short duration, 
always tied to the insecurity of political rule and never involved 
masses of the people except as temporary mobs (Hsu 1953:246- 
248) . The case of Hindu-Moslem violence and casteism in India is 
considered elsewhere (Hsu 1961) . Again, religious dissensions, per- 

* This is based on two groups of answers to a questionnaire. The first group of answers was 
from 230 Northwestern University students in 195 1 of whom 210 were undergraduates. A con- 
densed version of the same questionnaire was sent to a random sampling of 730 undergraduates 
at nine midwestern universities and colleges in 1952-53, from which 325 undergraduates re- 
sponded. The results, though quantitatively inconclusive, are qualitatively suggestive. First, 
students of Jewish background experience relatively little anti-Semitism at high school level when 
mixed dates are frequent, but at the university level their social contacts bcome much less 
diversified. Second, there is more open identification with Jewish culture and institution as the 
generation of Americanization advances. That is to say, the second and third generation Ameri- 
can Jews tend to be more openly Jewish than the fresh immigrants or first generation Ameri- 
cans. Coupled with this, Jewish students from families of higher social statuses (such as proprietary 
and professional) show more open identification than those from families of lower social statuses 
(such as sales). Third, in spite of these facts, students of Jewish background do not seem to 
prefer exclusive Jewish friendship and association in college. Fourth, with the term "normal adjust- 
ment" meaning acceptance by Gentile students, "the conscious (but not self-conscious) and 
self-identifying Jews among the students are those most integrated with their own people and 
the most normally adjusted on the college or university campus" (209). These results correspond 
amazingly to my personal observations but any final conclusion on the subject must, of course, 
await further research. 


secutions, and racial conflicts are today more intense and widespread 
in Protestant-dominated societies of the West (see Chapter 14) 
than in their Catholic counterparts. In this dichotomy we are con- 
trasting the United States, Canada, Australia, Union of South 
Africa, and so forth, as one camp and the Latin American repub- 
lics, as well as Portuguese, Belgian, and French African possessions 
as the other. What has happened in Protestant-dominated societies 
is that, by and large, persecution in the form of bloody racial and 
religious outbreaks has been consistently driven underground while 
the manifestations of prejudice have become diffused, one almost 
may say democratized if not for the fact that the expression smells 
of sarcasm. But even in the most advanced Protestant societies racial 
and religious violence is always around the corner, ready to erupt 
now and then, here and there, as indicated by the recent anti-Negro 
outbreaks in England and the recurrent anti-Semitic flare-ups in 
Europe and the United States.^ 

Three Uses of Value 

It will have been clear to some readers that this analysis of the 
psychosocial origin of racial and religious prejudices bears some re- 
semblance to that of Kurt Lewin on the problems of the Jews as a 
minority group in many a western society. But it has significant dif- 
ferences. According to Lewin the most basic problem of the Jews is 
that of group identity. Often repudiated in the country of his birth 
and upbringing, yet having no homeland which he can claim as his 
own, he suffers from "additional uncertainty," thus "giving" him 
"some quality of abnormality in the opinion of the surrounding 
groups." He concludes that the establishment of a Jewish homeland 
in Palestine (which was not yet a reality at the time of his writing) 
might "affect the situation of Jews everywhere in the direction of 
greater normality" (Kurt Lewin 1935:175—187). 

The Jewish minority certainly shares the central problem, with 
other minorities, of uncertainty of group identity. But our analysis 
also shows that the degree of this uncertainty depends, in the first 
place, on the basic value orientation of the host majority and, in the 
second place, on that of the minority groups themselves. There is, 
for example, every reason to expect the Jewish minority to have 
far less of a problem of identity in Latin American countries than 
in North American countries. As far as North America is con- 

° The place of Mohammedanism with reference to this analysis will be considered in another 


cerned, the Jews, like other minority groups, will always have the 
problem of identity whether or not they have a homeland. The 
Latin American peoples have less of the value orientation of self- 
reliance and, therefore, the individual has less psychosocial need to 
reject minority groups to maintain his status in society. On the 
other hand, within the United States, there is good reason to expect 
the Jewish minority to have a little more of a problem of identity 
than the Chinese and Japanese minorities even after the establish- 
ment of Israel. This is despite the fact that the Orientals possess 
much greater physical distinctiveness than the Jews as a whole from 
the Caucasoid majority. For the Chinese and Japanese have stronger 
ties with their families and wider kin groups than do the Jews, and 
are, therefore, less self-reliant and less free but more protected from 
the uncertainty of identity. 

In this chapter I have not differentiated the different uses to which 
the term value may be put. Charles Morris, in a book entitled Va- 
rieties of Human Yalue, postulated three such uses: "Operative" 
values refer to the "actual direction of preferential behavior toward 
one kind of object rather than another." "Conceived" values refers 
to the "preferential behavior directed by 'an anticipation or fore- 
sight of the outcome' of such behavior," and "involves preference 
for a symbolically indicated object." He illustrates this meaning of 
value by the example of the drug addict who firmly believes that it is 
better not to be a drug addict because "he anticipates the outcome 
of not using drugs." "Object" values refer not to the behavior pre- 
ferred in fact (operative value) or as symbolically desired (con- 
ceived value) but to what is preferable if the holder of the value is 
to achieve certain ends or objectives (1956:10-12) . 

While it is obvious that the three usages of the term "value" are 
not mutually exclusive and must influence each other, it is equally 
obvious that they are not hard to distinguish. Applying this scheme 
to the American scene we shall realize that self-reliance is an opera- 
tive value as well as a conceived value. It expresses itself in two direc- 
tions. In the positive direction it expresses itself as the emphasis on 
freedom, equality in economic and political opportunities for all, 
Puritan virtues. Christian love, and humanitarianism. These values 
are far more conceived than operative. On the negative side self- 
reliance expresses itself as the tendency toward racial prejudice, 
religious bigotry, laxity in sex mores, and totalitarianism. These 
values are far more operative than conceived. Values which are more 
conceived than operative are of great symbolic importance, and 


will be militantly defended by the people cherishing them. The less 
they live up to such conceived values the more they are likely to 
defend them, because their failures are associated with feelings of 
guilt. Values which are more operative than conceived are of great 
practical importance, and will be strenuously pursued by the people 
needing them. The more they have to act according to such opera- 
tive values, the less they will admit their reality, since their actions 
also lead to feelings of guilt. At one extreme we shall find men who 
will openly fight to guard these operative values most flagrantly. 
At the other extreme we shall find men who will practice them by 
devious means. Those who hold on to these operative values openly 
and those who do so by subterfuge will share one common charac- 
teristic: both will deny their actions are motivated by prejudice and 
Christian hate. They will both insist that their actions are based 
totally on other reasons. In the South one ubiquitous reason is states' 
rights. In the North a widespread reason is property value or fear 
of intermarriage. When the real operative values are divulged acci- 
dentally, as it were, by one of those who share them, the reaction of 
the rest will be resentment against the simpleton who spoke out of 
turn and angry denial of everything he disclosed. These mechanisms 
are repeated so often on so many occasions, including the most 
recent (1959-60) Deerfield and Park Forest, Illinois, outbursts, 
that they need no further illustration or elaboration. 

However, the ideas of equality, freedom, and Christian love in- 
evitably affect all Americans because they are values that are con- 
ceived more than operative. They might even be described as the 
conscience of the American society. That is why failure to live 
according to them or outright opposition to them will both lead to 
guilt, denial, and subterfuge. There are men and women who cham- 
pion the cause of the more conceived values just as those who 
desperately cling to and fight for the more operative values. The at- 
titude of both sides toward their respective values tends to turn the 
values they champion into object values. That is to say, the cham- 
pions of equality, freedom, and Christian love can consciously use 
their values as tools for their ends, just as the champions of prejudice, 
bigotry, and Christian hate can also consciously use their values as 
tools for their ends. 

In the hands of some politicians and all demagogues the relation- 
ship between these values and the objects they desire often becomes 
transparently clear and undisguisedly selfish. It has been suggested 
that Hitler's hate campaign against the Jews was a major secret of 


his power. It is not surprising, therefore, that in the recent (1959- 
60) Chicago area integration outbursts, as with similar scenes else- 
where before, the opponents to integration charged their adversaries 
for promoting integration as a means of wooing Negro votes. But 
the link between the more conceived American values and the more 
operative values is the core American value of self-reliance. The 
supporters of both desire social arrangements in which their own 
particular nests will be feathered in their own particular ways. 

As the emphasis on democratic equality and freedom and Chris- 
tian love increases with self-reliance, totalitarian racial prejudice 
and bigotry and Christian hate will also increase with it. When the 
individual is shorn of all permanent and reliable moorings among his 
fellowmen, his only security must come from personal success, per- 
sonal superiority, and personal triumph. Those who are fortunate 
enough to achieve success, superiority, and triumph will, of course, 
bask in the sunshine. To them democratic equality and freedom 
and Christian love are extremely laudable. But success, superiority, 
and triumph on the part of some must of necessity be based on the 
failure, inferiority, and defeat on the part of others. For the latter, 
and even for some of those who are in the process of struggling for 
success, superiority, and triumph, the resentment against and fear 
of failure, inferiority, and defeat must be widespread and often un- 
bearable. To them totalitarian prejudice and bigotry and Christian 
hate can be means to a flitting security. By pushing others down they 
at least achieve the illusion of personal success, personal superiority, 
and personal triumph.^ 

The Problem of Pessimism 

If the conclusions of this analysis seem to lend themselves to pes- 
simistic inferences, I wish to assure the readers that this is neither 
intentional nor desired. But the rule of science is that we must con- 
template whatever conclusions our evidences lead us to, whether 
they are pleasant or unpleasant. 

In attentuation of certain pessimistic notes in the conclusions 
reached we need, however, to realize that the contribution of West- 
ern self-reliance to human development has been great and that even 
the chains of conformity and organization have their salutary 

'Additional substantiation for this analysis is found in Carl J. Friedrick (ed.), Totali- 
tarianism, which contains the results of a conference of scholars in 1953 under the auspices 
of the American Academy of Arts and Sciences. Its conclusion is that totalitarianism is a new 
disease peculiar to modern culture. Modern culture here refers, of course, to Western culture. 


aspects. What gave the Western man his superiority over the rest of 
the world during the last 300 years was not his religion or his ro- 
manticism but his self-reliance and his competitive organization. 
It was his self-reliance which led him to discard the shackles of 
paternal authority, monarchical power, and medieval magic, in 
favor of wider organizations such as church and state, mercantile 
fleets, and industrial ventures. When the West met the East, it was 
the Western man's well-organized armed might which crushed the 
East. As late as 1949 one high-ranking United States official at- 
tributed civil war-torn China's plight, in a Harper's magazine ar- 
ticle, to the fact that the Chinese were "organizationally corrupt." 
It is instructive to note that today, the two giants of the West, the 
U.S.A. and U.S.S.R., are still most attractive to the rest of the world 
by their skill in organization. In various parts of the world their 
experts are helping peoples of other nations to organize their 
educational systems, or their marketing arrangements, or their 
agricultural practices, or their industrial efforts, or their military 
capabilities, or their national finances.' 

The purpose of this paper is neither optimistic nor pessimistic. It 
is to place the much-lauded American values in their proper genetic 
perspective. When this is done, we find that the best of America is 
directly linked with her worst, like Siamese twins. The way out of 
the worst is not to deny it but to recognize it for what it is. 

' The problem of why some individuals assume some aspects of the value orientation of their 
society more than other aspects is outside of the scope of this chapter. That problem is treated 
intensively in the works of Mering (1961), Kluckhohn and Strodbeck (1961), and others. 


Allport, Gordon 

1954 The nature of prejudice. Cambridge, Addison- Wesley Publishing Com- 
pany, Inc. 
Bain, Read 

1935 Our schizoid culture. Sociology and Social Research 19:266—276. 
Coleman, Lee 

1 94 1 "What is American: a study of alleged American traits. Social Forces, 
Vol. XIX, No. 4. 
CoMMAGER, Henry Steele 

1950 The American mind. New Haven, Yale University Press. 
Cuber, John F. and Robert A. Harper 

1948 Problems of American society: values in conflict. New York, Henry 
Holt & Co. 
Friedrick, Carl J. (ed.) 

1954 Totahtarianism. Cambridge, Mass., Harvard University Press. 


HoMANS, George C. 

1950 The human group. New York, Harcourt Brace & Co. 
Hsu, Francis L. K. 

1953 Americans and Chinese: two ways of hfe. New York, Abelard-Schuman, 

i960 Rugged individuahsm reconsidered. The Colorado Quarterly 9:143- 

196 1 Clan, caste and club: a comparative study of Chinese, Hindu, and 
American ways of life. Princeton, N.J., Van Nostrand Co. 
Kluckhohn, Clyde 

1941 The way of life. Kenyon Review, Spring, pp. 160—180. 
Kluckhohn, Clyde and Florence R. Kluckhohn 

1947 American culture: generalized orientation and class pattern, Chapter 
IX of Conflicts of power in modern culture, 1947 Symposium of Con- 
ference in Science, Philosophy and Religion, New York, Harper and 

Kluckhohn, Florence and Fred Strodbeck 

196 1 Variations in value-orientations. Evanston, 111., Row Peterson and Co. 
Laski, Harold J. 

1948 The American democracy. New York, The Viking Press. 
Lewin, Kurt 

1948 Psycho-sociological problems of a minority group. In Character and 
Personality, Vol. Ill, 1935, 175—187. (Reprinted in Kurt Lewin: Re- 
solving Social Conflicts, New York, Harper & Bros.) 
Mering, Otto Von 

196 1 A grammar of human values. Pittsburgh, University of Pittsburgh 
Morris, Charles 

1956 Varieties of human value. Chicago, University of Chicago Press. 
Myrdal, Gunnar 

1944 An American dilemma. New York, Harper & Bros. 
Nottingham, Elizabeth K. 

1954 Religion and society. New York, Doubleday & Co. 
Teitelbaum, Samuel 

1953 Patterns of adjustment among Jewish students. Northwestern Uni- 
versity, Ph.D., dissertation. 

Warner, Lloyd 

1953 American hfe: dream and reality. Chicago, University of Chicago Press. 

Williams, Robin M. 

195 1 American society, a sociological interpretation. New York, Alfred 
Knopf, (i960, 2d ed.). 

1956 Religion, value-orientations, and intergroup conflict. The Journal of 
Social Issues 12:14—15. 


Two PROJECTIVE instruments, the Rorschach and the Thematic 
Apperception Test, have practically become standard stock-in- 
trade of many anthropologists. However, the popularity of projec- 
tive tests in anthropological studies has waned greatly in the last few 
years. All sorts of objections and doubts have been raised about their 
cross-cultural validity, or their validity as an instrument of study- 
ing anything other than individual differences, or, among psych- 
ologists, even their validity for their original purpose of diagnosing 
individual maladjustment. 

Being the editor of two volumes containing Rorschach and TAT 
protocols collected by anthropologists from many parts of the 
world, and having intensively used the Rorschach as an instru- 
ment in the anthropological field situation, Kaplan is perhaps 
the most qualified psychologist to analyse the role of projective test- 
ing in psychological anthropology. Kaplan considers the merits 
or demerits of projective testing in psychological anthropology 
from two broad aspects. The first aspect is more general and con- 
cerns the efficiency of projective instruments in fathoming per- 
sonality. It is understood that no projective test, by itself, pretends 
to be a complete measure of personality. But there are indications 
that the Rorschach or any other single projective test may add lit- 
tle to the description of a personality beyond that provided by life 
history materials and observation studies. On the other hand, it is 
clear that, since all individuals must live as members of social groups, 
the most important things are not the total psychological charac- 
teristics of an individual but his functioning or "socially required" 
motivational characteristics. It is the latter characteristics which 
the tests, in conjunction with other sources of data, help to reveal, 
and which are the primary concern of the psychological anthro- 

The other aspect of Kaplan's chapter is concerned with several 
specific problems in the cross-cultural use of the Rorschach (and in 
a secondary way other tests) by psychological anthropologists. For 
example, the average number of Javanese responses to the Rorschach 



cards is very much larger than that of Thai responses, and so forth. 
How do we interpret such differences? Another problem Kaplan 
deals with is the possibly greater importance of the content of 
Rorschach responses than of the formal characteristics in the tra- 
ditional scoring procedures. 

Wallace's chapter will surely spearhead a renewed interest among 
anthropologists in the physical and biological factors in abnormal 
behavior. For many decades the anthropologist, like the psychia- 
trist, has tended to favor environmental rather than genetic deter- 
minants; and within the environmental, to favor almost exclusively 
social rather than physical determinants. Wallace spells out the phi- 
losophy underlying a new organic approach to mental illness and 
points the way to a possible synthesis between this and the func- 
tional approach which has dominated the psychosocial tradition in 
psychiatry and the social sciences. In this trail blazing effort, Wal- 
lace explores one of the well-known yet most puzzling of mental 
illnesses found among Polar Eskimos: Pibloktoq, sometimes trans- 
lated as arctic hysteria, with two alternative hypotheses — one based 
on calcium deficiency and the other on the Eskimo cultural pattern 
of withdrawal when the individual's confidence in his own ability 
to carry out a struggle is shaken. In the last part of his chapter Wal- 
lace constructs one of the most sophisticated models of the intricate 
relationships between type of organic illness, the victim and his so- 
ciety's responses to the illness, and the culture of the victim and his 

Man's attempt at reading dreams goes back as far as any cultural 
records, but his scientific understanding of the phenomenon is of 
very recent origin. Have dreams influenced the development of 
human thought? What do we know about universal symbolism in 
dreams? What are some of the interpretations of dreams in non- 
Western cultures? Are cultural differences correlated with differ- 
ences in types of dreams? What is the significance of the diverse 
attitudes and practices regarding dreams in different cultures? 
These are questions dealt with in D'Andrade's chapter on dreams. 
The foundation of modern scientific dream study remains Freudian 
in theory: that dreams reveal some motivational characteristics of 
the individual which are otherwise hidden, but the Freudian view 
that dream language is obscure is largely giving way to the view, 
based on much modern research, that the most important contents 
of dreams tend to be manifest. The last part of D'Andrade's chapter 
deals with the conditions affecting dream usages. Here he sum- 


marizes the results of his own cross-cultural study of the relation- 
ship between presence or absence of the anxiety about being alone 
and involvements with particular types of use of dreams. 

Campbell gives us a summary, from the point of view of a social 
psychologist, of two outstanding anthropological contributions to 
personality psychology: cultural relativism and the human "labo- 
ratory" situation. But the major purpose of Campbell's chapter is 
to outline some of the important methodological issues in anthro- 
pological research so far as he sees them, and to offer suggestions 
regarding alternative procedures. Some anthropologists may re- 
gard some of Campbell's indictments as being unfair. Some may 
object to some of Campbell's methodological suggestions on the 
ground that he wants us to fly before we can crawl. Others may not 
share to the same extent Campbell's enthusiasm for the Whiting and 
Child approach, represented in Chapter lo by D'Andrade and ex- 
plained in Chapter 1 2 by Whiting himself. But there is no doubt 
that Campbell has put his fingers on a number of methodological 
problems which sorely need systematic attention by those psycho- 
logical anthropologists who hope for greater scientific gains. 

chapter 8 



University of Kansas 

During the past decade a somewhat violent argument has arisen 
concerning the role of projective techniques in anthropological 
studies. Since the whole culture and personality area has somehow 
become prominently identified with these tests, it is of some impor- 
tance that the value and significance of the tests be assessed and that 
an understanding of their particular role and function be achieved. 
The use of projective tests in cross-cultural settings has flourished 
over the past two decades, and one may estimate that there have 
been as many as 1 50 studies in more than 75 societies. There appears 
to be sufficient work done so that the usefulness of the tests can be 
evaluated and their main difficulties and problems delineated. 

Since it is my belief that the ultimate judgment about the tests 
will be based on demonstrated utility or lack of it in relationship 
to the purposes of research workers, I shall attempt to make these 
purposes explicit, thereby specifying the theoretical and method- 
ological issues that projective test studies are relevant to. My plan 
is to discuss both the demonstrated values and difficulties of empiri- 
cal studies in relation to each of the purposes described. I shall in 
addition present my own position with respect to the kinds of per- 
sonality data required in the culture and personality field and dis- 
cuss the prospects of obtaining them by using projective techniques. 

The Delineation of Modal Personality Processes 

It is fairly clear that the great majority of cross-cultural projec- 
tive test studies are concerned with describing the personality char- 
acteristics that are most prevalent in particular cultural groups. The 
concept of "modal" or "basic" personality as introduced by Kardi- 



ner and Linton in the late 1930's has been perhaps the most influen- 
tial theoretical conception in the culture and personality field, and 
there has been widespread acceptance of the notion that in each cul- 
ture there exists a core of personality characteristics which are 
found in most members of the group. Until relatively recently the 
existence of this core of homogeneity has been regarded almost as 
axiomatic, and it has seemed very natural to culture and personality 
workers to begin with the idea that they should describe these typi- 
cal or "modal" characteristics. That this aim prejudges an empirical 
issue which has perhaps not been adequately settled, has not seemed 
to trouble the scientific conscience of culture and personality work- 
ers. At the present time there is beginning to be more respect for the 
variability that exists within societies which has, whenever it was 
studied, been found to be embarrassingly large (Inkeles and Levin- 
son 1956, Kaplan 1954, Wallace 1952, Vogt 195 1, DuBois 1944), 
and it even seems respectable to voice a doubt about the existence of 
modal characteristics. (See Adcock and Ritchie's factor analytic 
study, 1958, which found that all of the Rorschach differences be- 
tween groups of white and Maori subjects could be explained by 
one factor, imaginative thinking. However, these writers appear to 
attribute the paucity of differences to the failure of the Rorschach 
test rather than to accept the findings of the test.) 

In part the prejudgment of the issue of homogeneity has arisen 
from the failure to make explicit distinctions between the two main 
concepts — culture and personality. The most popular and widely 
accepted opinions (Spiro 195 1, Smith 1954) tended to obscure the 
distinctions between these terms. Some (Kluckhohn, for example) 
have asserted that culture and personality are simply abstractions 
from the same behavior and have used such phrases as "culture in 
personality" or "personality in culture." When modal personality 
is regarded as synonymous with learned cultural behavior, there can 
be no question about its existence since the very concept of culture 
implies the existence of uniformities and regularities. 

Projective techniques have fit the purposes of workers attempt- 
ing to describe modal personality processes. The Rorschach has been 
particularly easy to work with since no matter what subjects did 
with the test, responses could be scored and scores averaged to get 
measures of central tendency. When such averages are unaccom- 
panied by measures of variability, they are worse than worthless 
since they have left the worker satisfied and pleased with his errors. 
Unfortunately the addition of measures of variance complicates 


the situation since there is no standard criterion which will tell the 
worker when his group is homogeneous enough to be characterized 
vahdly by the mean or mode. It is perhaps not necessary to comment 
at this late date on the dubious practice of pooling Rorschach scores 
of individuals to arrive at a combined psychogram which is then 
taken to represent the group modal pattern. This yields a very tidy 
result but unfortunately one that very often seems to have no re- 
lationship to the patterns that are found in any of the individuals 
in the group. The derived pattern is a completely synthetic one, and 
the fact that such patterns have been found to be related to cultural 
factors is testimony to the ingenuity of research workers in being 
able to find relationships between almost any variables under the 

Wallace has made a serious and sophisticated attempt to deal with 
the problem of deriving a modal pattern ( 1952) . In analyzing Ror- 
schach records of the Tucarora group, Wallace computed the modal 
score for each of 2 1 scoring variables and then set confidence limits 
of 2 S.D.'s around each mode. He defined responses that fell within 
these limits as members of the modal class and then asserted that 
subjects whose scores fell within these limits on all 2 1 variables were 
members of the modal group. He found that 37 per cent of the 
Tuscarora were in this group while only 5 per cent of the Ojibwa 
were. Our admiration for the ingenuity of this attempt to develop 
some basis for defining modality is perhaps qualified by the arbi- 
trariness of the limits that were set, and one is left with the question 
of whether Wallace's modal class is too large or too small or is noth- 
ing more than a statistical accident. 

The existence of wide variability is not an insurmountable ob- 
stacle to modal personality analysis. It is merely necessary to follow 
the derivations of the modal picture with a second step which checks 
back to assess the applicability of the modal picture to each of the 
individuals in the group in something of the manner of Wallace's 
study. While this may appear to be an unwelcome complication and 
an addition to the labor of the research, it does seem fair to say that 
in the absence of some such back checking, validity cannot be 
claimed for one's conclusions. 

Projective tests do have a particular appropriateness to the task 
of drawing up modal personality pictures. In using them one ap- 
proaches the task in the simplest and most direct way possible. A 
series of individuals, hopefully a representative and unbiased sample 
of the population to which one wishes to generalize, is studied one 


by one by means of a standardized procedure. The question is asked: 
what is this person hke? And the answer, uncontaminated, it is 
hoped, by knowledge of other individuals or by the expectations 
one has about the findings, is based upon concrete and more or less 
standardized interpretations of specific pieces of information. The 
tests allow one to reduce what is obviously a task of great complexity 
and difficulty to manageable proportions. It is perhaps not too much 
to say that culture-and-personality study could not proceed with- 
out these or equivalent techniques. And, the availability of this 
method has resulted in a plethora of research. 

Nevertheless, one sometimes has the feeling that the problem has 
been made deceptively simple. The first assumption, for example, 
that the projective test samples adequately the personality processes 
of the individual in whom one is interested is an extremely hazard- 
ous one. There is ample reason to believe from intensive studies of 
individuals using many techniques (Murray 1938) , or from studies 
of successive administrations of a single test (Kaplan and Bergei 
1956) , that the data obtained from a single test is little more than a 
fragment which may on occasion have some central importance but 
which at best is only part of the story of personality. A single Ror- 
schach, TAT or both, even when augmented with life history ma- 
terials and extensive observation studies such as in Vogt's (1949) , 
must yield an incomplete account of the person. To the extent that 
the anthropologist or psychologist believes that personality is en- 
capsulated in the microcosm of the test protocol, he is undoubtedly 
in error and in particularly serious error because he isn't likely to be 
aware of it. When the protocols are sparse and inexpressive as is 
sometimes the case, it is even more foolish to believe that one has the 
truth or some substantial portion of it. When the worker knows 
that his sample is seriously incomplete, as most psychologists do, but 
treats it as though it were not, he is equally in error. 

In addition to the "sampling error" in personality study, there is 
the very difficult problem of interpretation. Characteristically, pro- 
jective techniques yield very interesting but somxewhat cryptic re- 
sponses. These responses, whether they afe Rorschach responses or 
TAT stories, are difficult to interpret even under the best conditions. 
When a drastic cultural difference exists, which the interpreter be- 
cause of his inadequate knowledge of the culture and language of 
the subject cannot take into account, the responses are often com- 
pletely uninterpretable. Unfortunately in this situation where the 
interpreter has had the choice of either admitting his helplessness 


or of going ahead and making the interpretation he would if the re- 
sponse had been given by someone in his own culture, the latter 
course has been followed. The result has been that modal personality 
pictures have often had little more meaning than fairy tales. 

Perhaps the most usual solution to the problem of Rorschach in- 
terpretation has been to apply one of the standard systems for scor- 
ing responses and for interpreting the scores, the Klopfer and Beck 
systems being the main ones in use. Since these scoring systems make 
it possible to score any response whatever its origin, it becomes pos- 
sible, once the score has been obtained, to ignore the exotic nature 
of the response and average, summarize, and interpret the scores 
in the usual manner. Quite aside from the doubtful validity of ap- 
plying these interpretative categories cross-culturally, it appears 
that Rorschach practice has, in the past decade, been swinging 
slowly away from preoccupation with scoring, toward an interest 
in the content of the responses. Formerly, content analysis was 
treated as an adjunct to the interpretation of the scores. Today, 
however, the situation is reversed and the principal approach is 
usually to an understanding of the expressive imagery of each re- 
sponse; the pattern and sequence of such expressions enable the 
worker to form a picture of some aspects of the emotional life of 
the subject. George DeVos (1952, 1955) has been the principal 
exponant of content analysis in cross-cultural work with the Ror- 
schach. His system for analyzing and scoring the content of the 
responses provides a welcome technique for summarizing the emo- 
tional imagery of the responses in a relatively objective way, and 
makes possible at least rudimentary quantative analysis of differ- 
ences between groups. 

It is unfortunate that one cannot have greater confidence in the 
more usual scoring categories. They provide very tempting mate- 
rial for numerical manipulation. The scores of persons in a particu- 
lar group can be averaged, for example, and the averages for each 
category can be taken as a pattern representative of the whole group. 
Other techniques, such as range, the various indices of variability, 
the analysis of variance, and the various correlational techniques, all 
yield measures that are directly relevant to questions that culture 
and personality workers are concerned with, such as degree of vari- 
ability or homogeneity within groups or subgroups, differences be- 
tween groups and between subgroups, the identification of factors 
accounting for variability or homogeneity, the relationship between 
scores in different groups, and many others. However, until it is pos- 


sible to apply these techniques to rehable scores which are vaKd 
measures of what they purport to measure, it seems wisest to forego 
their use, especially since their spurious exactness may lead us to be 
content with findings of uncertain correctness. 

In the face of these difficulties which probably apply to some ex- 
tent to all modal personality studies it is obvious that research to 
date must be treated with great tentativeness. Whether it is better 
to proceed doing the best one can with the limited capacity one has 
or to retreat from tasks that are too formidable is difficult to say. 
In either case, the proper attitude is one which makes very limited 
claims and is explicitly aware of limitations. 

My judgment thus far has probably been somewhat harsh and 
has overlooked the positive values in these studies. I have suggested 
that most if not all modal personality studies utilizing projective 
techniques have most probably been arriving at incorrect descrip- 
tions of the people they are concerned with. However, it might 
equally well be suggested that they have also been correct in some 
part of their descriptions. While the shotgun approach in which 
some of the wildly fired shots hit their mark is not a method to 
be advocated, it will yield a slow accretion of sound facts if the un- 
sound ones are gradually culled out in the course of repeated studies. 
But it is very likely that much more than this is achieved. Many, of 
the descriptions are coherent with impressions gained otherwise, in 
fact so much to the mark that anthropologists who have used pro- 
jective tests have come to feel considerable confidence in them (see 
Lessa and Spiegelman 1954, Gladwin and Sarason 1953, and DuBois 
1944) . The tests apparently do provide descriptions which in part 
at any rate satisfy the needs of anthropologists for deeper insights 
into the people they are studying and further have a certain amount 
of congruence with materials derived independently. 

The Delineation of Cross-Cultural Differences in Personality 

Closely related to the assumption that modal personality patterns 
actually do exist is the belief that peoples in different cultures vary 
considerably from each other in their personality characteristics. 
While this belief has been so strong that most workers have not felt 
it necessary to study the matter empirically, there are a few studies 
which have utilized projective techniques in relation to it. Projec- 
tive tests offer a particularly sound approach to this question since 
they provide relatively standardized stimuli to which the reactions 


of different peoples may be compared. Perhaps the study which 
most directly treats the question of cross-cultural differences is my 
own (Kaplan 1954) . 

In this study Rorschachs from 170 young men in four cultures, 
Zuni, Navaho, Spanish American, and Mormon, were compared 
with respect to fourteen variables. It was found that there was sta- 
tistically significant variability from culture to culture in five of 
the fourteen variables. However, it was noted that within each of 
the four groups the variability was very great and this coupled with 
the fact that the between-group differences were smaller than ex- 
pected, led to the conclusion that there is less variability among cul- 
tures than was expected. It should be noted that a severe limitation 
of this study is that it never went beyond the scores to consider the 
psychological traits that are presumed to underlie them. A number 
of other studies have done this in the context of descriptions of 
cross-cultural differences. Such studies include those of Billig, Gellin 
and Davidson (1947, 1948), Abel and Hsu (1949), Joseph and 
Murray's comparison of Chamorro and Carolinian Rorschachs 
( 195 1 ) , Strauss and Strauss' comparison of Sinhalese and American 
children's Rorschachs (1956-57), and Hsu, Watrous, and Lord's 
comparison of Hawaiian Chinese adolescents and Chicago White 
adolescents (1961).^ Each of these studies noted some differences 
between the groups they studied but also a great many variables in 
which no differences were noted. Although comparative studies 
have been relatively sparse, my general impression is that the pro- 
jective test is a useful if not essential technique and that further ex- 
plorations of this method are warranted. The difficult problem of 
interpreting the significance of various differences remains with 
us, but the method, by pointing out differences, can indicate many 
interesting and significant problems. 

During the past three years the publication in large quantities of 
the original protocols of Rorschach and TAT studies in Primary 
Records in Culture and Personality (Kaplan (ed.) 1956, 1957), 
a Microcard publication, has made possible the conduct of large- 
scale comparative studies for the first time. With the publication 
of Volume III of this series in 1961, the raw data of more than 65 
studies will be available and studies comparing Rorschach responses 
in 20 societies will be possible. To date more than 12,000 pages of 
personality materials have appeared. 

Henry and Spiro (1953) made a complete survey of such studies up to 1952. 


The Analysis of the Role of Personality Processes in 
Societal Functioning 

Under this heading we consider what is perhaps one of the two 
most theoretically and scientifically significant contexts in which 
projective tests are used cross-culturally, the other being the influ- 
ence of cultural factors on personality functioning. This problem 
is subsumed by the framework of the social scientist which is con- 
cerned with understanding the bases of social order and integration. 
An outstanding hypothesis which has been put forward by such 
eminent sociologists as Weber, Parsons, Merton, Fromm, and Ries- 
man holds that the motivational processes of individuals play a key 
role in societal functioning, the role having to do with the motiva- 
tion of socially required performances. In this connection, Inkeles 
andLevinson (1954) have urged that a distinction be made between 
the actual modal personality patterns that are empirically deter- 
mined to exist in members of a society and the "socially required" 
personality patterns that are needed for optimal societal function- 
ing. The latter consists of the core of motivations which lead indi- 
viduals to perform the socially necessary jobs and act in appropriate 

Until the last few years systematic distinctions between these two 
concepts had not been made and in most theoretical schemes the 
modal personality model did service as the socially appropriate char- 
acter structure also. At the present time a number of writers includ- 
ing Riesman, Parsons, Spiro, Devereux, Singer, Inkeles, and others 
have suggested that the appropriate social behavior is not a function 
of the total personality pattern of individuals but of particular and 
specific motivational structures. Consequently the problem is to 
describe these specially relevant characteristics rather than to de- 
scribe personality characteristics or motivations which interfere 
with social functioning. The remainder, which neither facilitate 
nor interfere with society, can from the point of view of this prob- 
lem be eliminated from consideration. 

A second point that might be made is that, from the point 
of view of society, the crucial matter is, as Parsons and Shils 
(1951:158) have stated "to get the patterns (of behavior) what- 
ever their functional significance to the person it does not matter 

whether there are important differences among types of personality 
possessing this need-disposition (to behave in the required way) as 
long as it exists." 


Thus it appears that it is not necessary to posit shared motiva- 
tional orientations or modal personality characteristics in order to 
account for social behavior. The motives of individuals may vary 
considerably. The important matter is that the jobs get done. 

What is the role of the empirical investigation of personality in 
this theoretical problem? It is clear that it is not simply the study of 
personality. Perhaps the best entering wedge is through the concept 
of the conformity-deviance dimension; that is, to define the prob- 
lem as having to do with the discovery of the actual motivational 
bases which lead to conf ormative behavior. While from one point of 
view it may not seem important to have any knowledge of what 
these motivational supports are, since obviously they exist or the 
society would not function at all, from another viewpoint it seems 
probable that problems of culture change, inadequate role perform- 
ance and of deviance, to mention only a few, all demand a knowl- 
edge of what the motivations of the person are relative to the 
demands that are made on him for social behavior. 

In another paper (Kaplan 1957) considering what some of these 
motivations might be, I have suggested that these bases do not lie 
in point-for-point isomorphism with specific social requirements 
and values, so that, for example, competitive behavior is supported 
by motives toward competition, since these can be understood as 
being primarily instrumental in nature and in a means-end relation- 
ship to other motives of a more generalized nature. We have posited 
instead the view that what we must look for are the generalized 
dispositions which are involved in the total relationship of the per- 
son to the social reality in which he acts, a reality which is to a con- 
siderable extent organized in normative terms — that is, which 
specifies what he should be doing and how. These generalized dis- 
positions are probably of a few main types which are widely dis- 
tributed. Riesman and Fromm especially have been concerned with 
the nature of these generalized dispositions and term them "social 
character." "Other directedness" is a perfect example of the kind 
of generalized disposition we are talking about, and Riesman has 
given a very rich account of the consequences that flow from a so- 
ciety's general reliance upon this type of motivational orientation. 
His discussion indicates that what is at stake is the very core and 
basis of social integration, and the essence of social order itself. 

The description of the dominant type of motivational orienta- 
tion is an empirical matter, and in line with a bias toward individual 
study in matters in which traits are ascribed to individuals, I believe 


that the most vaHd approach to this task is through the study of 
individuals. Riesman's Faces in the Crowd (1952) provides this 
kind of study but it is perhaps more of an illustration of his theory 
than an analysis of the facts. Can projective technique studies supply 
us with relevant information? My belief that it can is based on the 
following theoretical analysis. 

If we choose to understand personality processes as social action, 
that is, as being in the realm of what the person is doing rather than 
as something he has, the pattern which is established in a projective 
test protocol becomes an act or series of them which bears explain- 
ing. My view of action is that it is a function of a social reality which 
is organized around certain normative components and a motiva- 
tional orientation relative to this reality. If we regard the projective 
test protocol as a personality pattern which the person establishes 
for the moment through his action, the two main kinds of informa- 
tion that we require to understand or explain it are the delineation 
of the normative aspects of the situation which define the legitimate 
expectations that are perceived by the person and the motivational 
orientation which prescribes the position or stance which is taken 
relative to these expectations. The first is a matter for the social 
scientist since it has to do with the character of the social situation, 
while the second is a problem for the psychologist whose main in- 
terest is or should be in the analysis of motivation. Each response 
in the projective test can be analyzed from these two points of view. 
This, of course, places a very heavy burden on the test analyst and 
perhaps it will appear to the reader that the task is too difficult or 
impossible. The gains to the social scientist, however, are very large 
since they involve nothing less than an understanding of the rela- 
tionship of the actor to the phenomenal reality in which he exists. 

Although there probably has been little or no interpretation of 
projective techniques in this vein to date, there has been a certain 
amount of work that is concerned with the relationship of modal 
personality processes to the functioning of social patterns. Inkeles 
and Levinson (1954) discuss the problem of "congruence" be- 
tween modal personality traits and social requirements. Inkeles, 
Hanf mann, and Beier (1959) are concerned especially, in studying 
the Soviet social system, with determining the fit between the per- 
sonality traits that were determined in an extensive personality 
study utilizing projective techniques, principally the Sentence 
Completion Test, and the present requirements of the Soviet system. 
They suggest that there is a considerable degree of noncongruence, 
almost inevitable in any rapidly changing society, that has serious 


consequences for the functioning of the Soviet system. Dicks' 
(1952) study of Soviet personahty, although not utihzing projec- 
tive techniques, is in a similar vein. Erickson (1950) is another 
writer who has been concerned with this kind of problem. 

While we have discussed the problem of determining the moti- 
vational characteristics that are involved in conf ormative behavior, 
similar questions might be asked with respect to deviance. What 
is the nature of the motivational orientation which leads to a nega- 
tive relationship to the normative aspects of a situation? Again this 
is an empirical matter of the greatest importance which should be 
studied with projective techniques and all other available methods. 
The method suggested above in relation to the problem of conform- 
ity, of viewing the response or story as an action which stems from 
a particular relationship of the person to the normative, can be ap- 
plied in the analysis of deviance as well. In one sense a considerable 
part of the problem of action lies in the necessity of making a choice 
between the deviant and conformative alternatives that are present 
in the situation. 

Something might be said about the merits of the Rorschach and 
TAT relative to this problem. For any attempt to describe and 
analyze motivational processes it would appear that the TAT has 
important advantages. As the test has been used by Murray and his 
associates and by most other psychologists, the primary focus has 
been on describing the hierarchy and patterning of motives and 
their relationship to the perceived social environment. Since the 
stories are ordinarily comprised primarily of actions rather than 
descriptions of qualities or feelings, and psychologists have held that 
motives are inferable from actions, this instrument seems especially 
pertinent to the requirements of the worker in psychological an- 
thropology. The Rorschach, on the other hand, ordinarily provides a 
series of highly condensed and often cryptic visual images from 
which motivations are only indirectly inferable. One has the im- 
pression that these images pertain more to the cognitive organiza- 
tion of the emotional life than to the motivational or volitional 
elements. However, this is not the exclusive focus but rather one 
that is relatively stronger. 

Learning More about Projective Techniques Themselves through 
Cross-Cultural Studies 

One of the main motivations for many cross-cultural studies has 
been simply to see what Rorschachs or TATs or other tests look like 
in exotic societies. This has involved a mixture of wanting to see 


more of what the people are hke and wanting to see what the Ror- 
schachs of people who differ so much from ourselves would look 
like. Thus many sets of test protocols have been presented almost as 
fascinating curiosities not necessarily having any great scientific 

One aspect of this curiosity is practical. Clinical psychologists 
are continually faced with the problem of cultural diversity in their 
subjects. Class, ethnic, and regional differences are an ever-present 
part of the situation in which they work. While these factors are for 
the most part ignored, there is an uneasy feeling that they are sig- 
nificant and something should be done about them. Feeding into 
this is the uneasiness about the norms in terms of which tests are 
ordinarily interpreted. Rorschach workers have fairly clear and 
well-established ideas about normal performances and their inter- 
pretations are generally made in relationship to these norms. Sub- 
cultural differences in subjects raise a question about the general 
applicability of these norms. In this situation the psychologist looks 
toward the worker in psychological anthropology for guidance and 
help in establishing the importance of cultural factors for his own 
interpretations and for clarifying the ways in which these factors 
can be taken into consideration. 

Studies in perhaps seventy-five societies have not, unfortunately, 
served to settle these questions. The finding has been that societies 
do vary considerably in the typical performances that are given by 
their members, ranging from the sparseness and brevity of the 
Ojibwa Rorschachs, for example, to the richness and expressiveness 
of the Algerian, Japanese, and Hindu records and the cryptic and 
almost impossible to interpret records of the Melanesian peoples. 
The TAT records range from the two and three sentence records 
of Navaho and Hopi children to the fifty and seventy-five pages 
given by Javanese young men. What is perhaps most obvious is that 
the way the test works varies considerably from group to group. A 
large and important question is whether these differences result 
simply from the subject's approaching the test in a different frame- 
work and with different cultural conventions or whether they re- 
flect genuine differences in personality processes. The difficulty of 
separating these two possibilities is one of the chief obstacles of Ror- 
schach interpretation in cross-cultural settings. 

The discovery of this great variability which clearly transcends 
the individual variability that exists in our own society has been of 
great interest, however, to the extent that it reveals to us new modes 
of reaction and presents us with concrete examples of personality 


functioning which have radically different bases than exist in West- 
ern society. The understanding and appreciation of these differences 
should widen our understanding of the human species and of the 
possibilities that are open to it. 

What is the Best Way to Study Personality 

It is clear that projective tests "work better" in some societies 
than others in the sense that in some groups they yield more ex- 
tensive and richer information. This seems to be analagous to the 
fact that the tests work better with some individuals than with 
others. A general problem thus raised, might be phrased, "How do 
the characteristics of the people being studied influence the way 
that they should be studied?" 

The TAT study of Lucien Hanks, Jr. (1956) utilizing a set of 
specially drawn pictures paralleling the Murray cards, is most no- 
table for the sparseness of the stories told. The subjects, who were 
mostly agricultural workers from Bang Chan in Thailand, gave 
almost no fantasy material. Hanks, in trying to account for the 
briefness of the stories, raises the interesting question of whether 
the test situation has created inhibiting anxiety in his subjects or 
whether the ability to fantasy was undeveloped in his subjects. An 
examination of the records suggests that the key to their sparseness 
lies in the fact that the subjects, without exception, were not telling 
stories but simply describing what seemed to them to be happening 
in the pictures at the moment. While it seems possible that the Thai 
cannot tell stories or are reluctant to do so; that they did not un- 
derstand what was required of them or, understanding, did not 
know how or did not wish to comply, it is clear that both cognitive 
orientation and motivation are essential factors in the projective 
test situation and that without understanding them, it is almost 
hopeless to attempt an interpretation of materials from societies 
other than our own. While these factors can frequently be inferred 
from the records themselves, more often they remain unclear, and 
we are uncertain whether the variation of the records from those 
of some other culture is the result of differences in what the subjects 
were trying to do, differences in how hard they were trying, or ac- 
tual differences in the personality and intellectual characteristics 
of the subjects. Of course, these three factors are not independent 
since what we mean by personality characteristics is sometimes only 
that the subject prefers to do one thing rather than something else 
or that he has a tendency to understand things in some particular 


It is interesting to note, in line with this same research and in 
vivid contrast to the Thai's performances, the fantastically lengthy 
and intricate stories collected from young men in Java by Hildred 
Geertz (1957). Using a specially constructed set of TAT cards 
and recording the stories in the native language, Geertz obtained 
protocols averaging over fifty typewritten pages in length. It seems 
very possible that contrary to Hanks' findings, the Javanese have 
the capacity to give very rich, imaginative, and revealing fan- 
tasies. These conditions undoubtedly vary from culture to culture 
and from individual to individual. Their discovery requires a high 
degree of ingenuity and flexibility from the test administration, and 
an acknowledgment that the standard instructions and testing situ- 
ation must sometimes be abandoned in the search for the better 
conditions under which it is possible to elicit significant personality 
data. It also requires that the tester investigate the subject's under- 
standing of the test situation and the nature of his motivations and 
concerns about the test. Since this kind of research has hardly been 
done in our own culture, it is perhaps optimistic to expect that it 
can be done in cross-cultural studies. However, it does seem to be 
the very minimum needed. 

The general principle which should hold for all projective tech- 
nique interpretation is that the absence of some particular kind of 
material should not be regarded as indicating the absence of the 
ability necessary to produce the material. Instead, one should inter- 
pret what has been given as the preferred style or mode of the sub- 
ject under the particular circumstances of the existing situation. 
This principle is specially important in cross-cultural studies. The 
Rorschach study of Carstairs (1956) is very relevant to this issue. 
His extensive series was collected in Delwara village in Udaipur and 
in the Bhil tribe, also in Udaipur. Despite the fact that most subjects 
in the Delwara group were unsure of themselves and showed many 
signs of anxiety, these records are rich and interesting. Although a 
great many of the subjects seemed reluctant and anxious and felt 
that they were not doing what was required of them, they were 
appropriately oriented toward the task and gave the kinds of ex- 
pressive responses which Rorschach workers expect and hope for. 
Perhaps something in the cultural situation and in the "modal per- 
sonality" characteristics of the group is appropriate to the require- 
ments of the Rorschach test. 

Carstairs' Bhil Rorschachs were collected under much more fa- 
vorable circumstances than were the Hindu records. He reports that 
the Bhils seemed to enjoy the test and had an easy relationship with 


the tester. Despite this, they had a much more difficult time in giving 
responses and the responses are much less revealing. There is more 
stereotypy, vagueness, and rejection of cards. The content seems less 
emotionally charged and less symbolic in nature. A comparison of 
the two sets of Rorschachs suggests that far deeper and more pro- 
found factors than the immediate test situation are involved in the 
differences between the Bhil and Hindu records. The former, despite 
a great readiness to respond freely and spontaneously, gave com- 
paratively little; while the latter, despite considerable reticence, 
caution, and anxiety, were extremely expressive. It is difficult to say 
why this is so. One might speculate that two different kinds of ac- 
tions are involved: the personality of the subjects and their char- 
acteristic modes of cognition. It is not possible to specify at this 
time how these factors operate in the Hindu and Bhil groups. Con- 
ceivably, however, the greater anxiety and involvement of the 
Hindu group could stem from unsolved personality problems which 
were being worked out very near to the surface of consciousness. If 
this were the case, the Rorschach situation might have greater func- 
tional significance for the subjects. The Bhils, on the other hand, 
whose anxieties and problems apparently are considerably more re- 
pressed, did not find the situation of psychological use since they 
were not "working through" their problems, but were suppressing 

A number of studies suggest that the acculturation variable has 
something to do with expressiveness on projective tests, the general 
finding being (see Hallowell 1942, Spindler 1955) that accultura- 
tion is associated with greater expressiveness. An obvious point is 
that as nonliterate peoples become more and more influenced by 
western culture and become more like the population for whom the 
tests were devised, the tests will work better, in the sense of yielding 
richer and more valid data. As has been suggested above, the diffi- 
culty in using projective techniques cross-culturally is not only a 
matter of increased uncertainty about the validity of the tests but 
involves the sparseness of some of the materials and the inability to 
obtain rich, imaginative, personal, and expressive data in contrast to 
brief, superficial, and stereotyped responses with a minimum of 
personal involvement. Although the latter are certainly not un- 
known in our society and in some parts of the population may even 
predominate, it does appear that the Rorschach and TAT generally 
do yield better materials in our own society than in most others. The 
reasons for this are not completely obvious, especially where the 
Rorschach test is concerned. The success of these tests in our own 


society seems not to be based on an explicit recognition of the fea- 
tures of our culture and its people that make us more permeable to 
particular kinds of personality study procedures, but rather to be 
based either on pragmatic grounds or an intuitive understanding of 
what is appropriate in a personality study. 

One factor of importance to personality study is an openness and 
willingness to be known by others, the exact opposite of what 
Lerner (1961) finds in the French, who speak of the "refus de 
s'engager," who answer the telephone by saying, "Je vous ecoute," 
and who answer the greeting on the street, "Comment va?" with 
the ironic reply, "On se defend." Western society is not the only one 
in which this openness to personality study is found. Geertz's 
lengthy Javanese TATs and many other sets of data indicate that 
the quality of openness is widely distributed. However, it would 
probably be premature to state that openness to personality study 
is a general quality of any people. 

I have conducted informal studies with a group of young Navaho 
men which make it clear that certain techniques yield more infor- 
mation about personality in this group than others. These very shy, 
noncommunicative individuals proved to be very difficult subjects 
despite their apparent eagerness to co-operate and be of help. Ror- 
schach responses and TAT stories were sparse and unrevealing. On 
the other hand, the Rosenzweig Picture Frustration test yielded 
very good data. Perhaps most interesting were my attempts to get 
life history materials. Individual after individual gave the briefest 
and most impersonal possible account of his life. In varying the con- 
ditions of the study in an attempt to get more expressive materials, 
I found that if the subjects were allowed to write their life stories, 
they furnished quite lengthy and expressive accounts, despite 
considerable difficulty with pencil and written English. It seems, 
therefore, that two different problems exist, one having to do 
with the reasons for the differences in the general tendency to per- 
meability in different societies, and the other dealing with the varia- 
tions in the conditions under which individuals in different societies 
are willing and able to be personally expressive. 

With respect to the first of these problems, we might speculate 
that conceptions of personality and individuality prevalent in the 
culture are among the relevant factors. For example, in a culture in 
which there is considerable concern about self and where thought 
about differentiated individuality is high, Rorschach materials may 
be specially rich and revealing. Hallowell's (1954) analysis of con- 
cepts of self and of kinds of self-awareness as cultural variables 


influencing persons' self-images and experience of self is very rele- 
vant to this problem and offers many exciting leads. 

What Has Cross-Culturol Use of Projective Techniques 
Tauglit Us about Personality Development 

One of the great hopes and aims of cross-cultural personality 
study has been the feeling that a better understanding of personality 
functioning itself might be achieved if cultural factors could be 
given more serious consideration. It is difficult to say whether any- 
thing of real value has been accomplished along these lines. Perhaps 
the best criteria that can be utilized is whether any new conceptions 
of personality functioning have emerged as a result of these studies. 
Here my impression is that they have not. Of the theoretical work 
of anthropologists only A. I. Hallowell has made any significant 
contributions; his work on the self and some of his theoretical writ- 
ing being in my opinion of great importance for psychologists. 
However, while his work with projective techniques may have 
added to his psychological orientation and sophistication, it has in 
itself been of no great importance. 

Of the psychologists who have been influenced by anthropologi- 
cal work, Erik Erikson, Abram Kardiner, and Erich Fromm have 
made perhaps the only significant additions to our conception of 
personality functioning, the remainder of neo-Freudian social 
thinking coming fairly directly from Freud and Adler and not 
being related to postwar empirical work at all. One might say that 
psychoanalytic theory has had a much greater influence on the cul- 
ture and personality field than this field has had on psychoanalytic 
or other personality theory. Projective test studies have in the main 
been used to support and bolster conceptions which have emerged 
from these theories, principally the notion that child-rearing prac- 
tices have a crucial role in the development of adult personality 
characteristics. Considerable support for this hypothesis has been 
developed by empirical studies, although it is perhaps not com- 
pletely conclusive as yet. However, important influences of these 
studies on theories of personality functioning have not yet occurred. 

Nevertheless, there is perhaps some reason to be optimistic that 
such influence may not be too long in coming. Perhaps the fact that 
psychologists themselves have been very slow in coming into the 
culture-and-personality field is responsible. While anthropologists 
have shown a high degree of sophistication in the use of psychologi- 
cal concepts and a few like Hallowell have become first-rate psy- 
chological theorists, their contribution to what are essentially psy- 


etiological problems is necessarily limited. If, as perhaps can be 
anticipated in the not too distant future, psychologists in substan- 
tial numbers take up the problem, further theoretical development 
can be expected. 

A Summing Up 

My judgments about the cross-cultural use of projective tests 
have been very harsh. I have looked for the positive values in these 
tests and found them very scant. I have looked at the difficulties in 
their use and found them to be enormous, and have concluded that 
as these tests are being used and interpreted at present, only a modi- 
cum of validity and value can be obtained from them. 

Nevertheless, cross-cultural personality study is one of the most 
rewarding, exciting, and important areas in the social sciences. The 
diflSculties that have been noted are not in the least discouraging but 
on the contrary add to the feeling that this is an extremely produc- 
tive field which is just at the beginning of making a great contribu- 
tion to the development both of social and psychological theory. My 
criticisms of current practices have been aimed mostly at those who 
would suggest that the problems do not exist or can be ignored. If 
there is any general moral to my remarks, it is that psychological 
anthropology in the next decade must center around research in 
how to study personality and how to use these tools with depth and 


Abel, Theodora M. and Francis L. K. Hsu 

1949 Some aspects of personality of Chinese as revealed by the Rorschach test. 
Journal of Projective Techniques 13:28 5-3 o i . 
Adcock, C. J. and J. E. Richie 

1958 Intercultural use of Rorschach. American Anthropologist 60:881-892. 
BiLLiG, D., J. GiLLiN, and W. Davidson 

1947-48 Aspects of personaUty and culture in a Guatemalan community. Jovu-nal 
of Personality 16:153-187, 326-368. 

Carstairs, G. M. 

1956a Rorschachs of forty high caste Hindus and ten Moslem men from 
Delwara, Udaipus, India. In B. Kaplan (ed.) , Primary records in culture 
and personality. Madison, University of Wisconsin Press, Vol. i . 

Carstairs, G. M. 

1956b Rorschachs of twenty-two Bhil men from Udaipus, India. In B. Kaplan 
(ed.). Primary records in culture and personality. Madison, University 
of Wisconsin Press, Vol. i. 


DeVos, G. 

1952 A quantitative approach to affective symbolism in Rorschach responses. 
Journal of Projective Techniques 16:133-150. 

1955 A quantitative Rorschach assessment of maladjustment and rigidity 
in acculturating Japanese Americans. Genetic Psychology Monographs 

Dicks, H. V. 

1952 Observations on contemporary Russian behavior. Human Relations 

DuBois, C. 

1944 The people of Alor. Minneapolis, University of Minnesota Press. 
Erickson, E. H. 

1950 Childhood and society. New York, Norton. 
Geertz, H. 

1957 Modified TAT's of thirty-three Japanese men and women. In B. Kaplan 
(ed.), Primary records in culture and personaUty. Madison, University 
of Wisconsin Press, Vol. 2. 
Gladwin, T. and S. B. Sarason 

1953 Truk: man in paradise. New York, The Viking Fund. 
Hallowell, a. I. 

1942 Acculturation processes and personality changes as indicated by the 
Rorschach technique. Rorschach Research Exchange 6:42-50. 

1954 The self and its behavioral environment. Explorations, Vol. 2. Re- 
printed in A. I. Hallowell, Culture and experience. Philadelphia, Uni- 
versity of Pennsylvania Press, 1955. 

Hanks, L. 

1956 Modified TAT's of forty-seven Thai children and adults. In B. Kaplan 
(ed.). Primary records in culture and personality. Madison, University 
of Wisconsin Press, Vol. i. 

Henry, Jules and Melford E. Spiro 

1953 Psychological techniques: projective tests in field work. In A. L. 
Kroeber and Others, Anthropology today, pp. 417—429. Chicago, Uni- 
versity of Chicago Press. 
Hsu, Francis L. K., Blanche Watrous and Edith Lord 

1 96 1 Culture pattern and adolescent behavior. International Journal of Social 
Psychiatry, Vol. VII, No. i, 33-53. 
Inkeles, a., E. Hanfmann, and H. Beier 

1959 Modal personality and adjustment to the Soviet socio-political system. 
Human Relations, Vol. 2, No. i. Reprinted in B. Kaplan (ed.). Study- 
ing personality cross-culturally. Evanston, Row, Peterson, and Co., 
Inkeles, A. and D. J. Levinson 

1956 National character: the study of modal personality and socio-cultural 
systems. In G. Lindsey (ed.), pp. 977—1020. Handbook of social 
psychology. Cambridge, Addison-Wesley. 
Joseph, Alice and Veronica F. Murray 

195 1 Chamorros and Carolinians of Sapian: personality studies. Cambridge, 
Harvard University Press. 


Kaplan, B. 

1954 A study of Rorschach responses in four cultures. Papers of Peabody 
Museum of Archeology and Ethnology. Harvard University, Vol. 42, 
No. 2, 

1956-57 Primary records in culture and personality. Madison, University of 
Wisconsin Press, Vols. 1—2. 
1957 Personality and social structure. In J. B. Gittler, Review of sociology: 
analysis of a decade. New York, Wiley and Sons. 

Kaplan, B. and S. Berger 

1956 Increments and consistency of performance in four repeated Rorschach 
administrations. Journal of Projective Techniques, Vol. 20, No. 3. 

Lessa, W. a. and M. Spiegelman 

1954 Ulithian personality as seen through ethnological materials and thematic 
test analysis. University of California Publications in Culture and So- 
ciety. Berkeley and Los Angeles, University of California Press, Vol. 2. 

Lerner, D. 

1 96 1 Interviewing Frenchmen. In B. Kaplan (ed.). Studying personality 
cross-culturally. Evanston, Row, Peterson and Co. 

Murray, H. A. 

1938 Explorations in personality. New York, Oxford Press. 

Parsons, T. and E. A. Shils 

195 1 Toward a general theory of action. Cambridge, Harvard University 

Riesman, D. 

1952 Faces in the crowd. New Haven, Yale University Press. 
Smith, M. B. 

1954 Anthropology and psychology. In J. Gillin (ed.), pp. 32-66. For a 
science of social man. New York, Macmillan. 

Spindler, G. 

1955 Socio-cultural and psychological processes in Menomini acculturation. 
University of California Publications in Culture and Society. Berkeley 
and Los Angeles, University of California Press. 

Spiro, M. 

195 1 Culture and personality: the natural history of a false dichotomy. 
Psychiatry 15:19—46. 

Straus, M. A. and J. H. Straus 

1956-57 Personal insecurity and Sinhalese social structure: Rorschach evidence 
for primary school children. Eastern Anthropologist, Vol. 10, No. 2. 

VoGT, E. A. 

195 I Navaho veterans: a study of changing values. Papers of Peabody Mu- 
seum of Archeology and Ethnology. Harvard University, Vol. 4 1 , No. i . 

Wallace, A. F. C. 

1952 The modal personality structure of the Tuscarora Indians as revealed 
by the Rorschach test. Smithsonian Institute, Bureau of American 
Ethnology, Bulletin 150. 

chapter ^ 



The University of Pennsylvania 


Do DIFFERENT culturcs encourage different styles of mental illness? 
Are there societies in which mental illness is absent, or at least rare in 
comparison with our own? Have either style or frequency of men- 
tal illness, or both, changed during the history of Western civiliza- 
tion? These and similar questions, prompted by practical concern 
with the mental health of our contemporary world populations, 
have evoked answers from anthropologists. Yes, different cultures 
do encourage different styles of mental illness, but the major cate- 
gories of mental illness (the organic psychoses, the functional psy- 
choses, the neuroses, the situational reactions, etc.) seem to be 
universal human afflictions. No, there are no societies of whom it 
can be said with confidence that mental illness is absent or, with 
certainty, that it is even rare, but there are certainly differences in 
the frequencies of illness and in the readiness of different social sys- 
tems to recognize what Western psychiatry would call illness as 
significant disorder. Yes, styles and frequencies of various mental 
illnesses have changed in recent western history (hysteria, for in- 
stance, is now a relatively rare diagnosis, and devils and demons have 
been replaced by radio and radar in paranoid delusions) , but we do 
not know all of the reasons for such changes over time nor for the 
differences between social classes and between regions. 

Thus, the relation between culture and mental health remains an 
intriguing problem for anthropologists, a promising field for re- 
search, and perhaps some day a richly rewarding field for applica- 
tion. At the present time, like other scientists interested in mental 



illness, anthropologists are still searching for more adequate con- 
cepts, more powerful theories, and more effective techniques of 
observation. One of the avenues of research which has been under 
rapid construction outside of anthropology is biological in con- 
cept and method; and since this approach is relatively unexploited 
by anthropologists, yet is potentially of great significance for an- 
thropological theory, a considerable part of this chapter will be 
devoted to considering the ways in which the current cultural- 
anthropological work in this area can assimilate and exploit what 
may be regarded, in the context of anthropology, as a physical- 
anthropological position. 


The culture and personality tradition in anthropology has bor- 
rowed its models of personality development, its characterology, 
and its conceptions of mental illness almost exclusively from a com- 
bination of learning, Gestalt, and psychoanalytic theories. This is in 
part a historical accident: these functional approaches were de- 
veloping most vigorously in American psychology and psychiatry 
just at the time, in the late 1920's and early 193 o's, when cultural 
anthropologists were first turning their attention seriously to the 
individual. Anthropologists found these psychologies readily ap- 
plicable to an understanding of the individual in culture; and the 
psychologists and psychoanalysts found in cross-cultural materials 
useful corroborative evidence for their theories. But the more re- 
cently developed biological approach, while it has not as yet (any- 
more than the functional approach) provided a spectrum of 
"cures" of such refractory disease clusters as schizophrenia and 
cerebral arteriosclerosis, has already yielded a considerable body of 
knowledge of processes (in this case, of organic mechanisms) which 
are implicated in one or another type of psychopathology. This 
knowledge should be incorporated without delay, in general out- 
line, into the conceptual armamentarium of every anthropologist 
concerned not only with mental disease but also with normal per- 
sonality development and function. 

At the present time, anthropological treatments of mental disease 
topics, particularly by culture and personality scholars, generally 
depend on a simple paradigm: the symptomatology of the illness 
under scrutiny is assumed to be motivated behavior expressive of 


psychological conflicts and to some degree effective in reducing ten- 
sion and anxiety; the symptoms are "interpreted" in terms of some 
deductive schema intended to lay bare the (usually assumed to be 
unconscious) conflict; cultural Anlagen in the symptomatic be- 
havior are pointed out; and finally, the source of the conflict is 
sought in traumatic emotional and/or cognitive dilemmas imposed 
by the victim's culture. This procedure almost completely neglects 
the victim's body; or, rather, it attributes to the victim's psyche a 
virtually magical ability to control the state of its body, by un- 
critically assuming that almost any somatic expression can be satis- 
factorily explained merely by asserting a plausible concomitant 
intrapsychic conflict. Even the "psychosomatic" position, it must 
be emphasized, is not "organic" in the sense indicated above, for it 
seeks the explanation of both somatic and behavioral disorder in 
antecedent psychological and cultural rather than in antecedent 
physiological conditions: thus the ulcer is explained by reference 
to the autonomic discharge attendant upon intrapsychic conflict, 
and the existence of intrapsychic conflict is explained by reference 
to culturally enjoined learning experiences rather than by any neu- 
rophysiological process. 

Thus, even with regard to syndromes familiar to Western clini- 
cians and conventionally (if not invariably) conceived as func- 
tional in etiology, the assumption that biological determinants are 
negligible is becoming an increasingly hazardous one to make. But 
the anthropologist is peculiarly vulnerable to criticism when he 
utilizes the functional paradigm without qualification to explain 
exotic forms of mental illness, such as the pibloktoq of the Polar Es- 
kimo and the windigo psychosis of the northern Algonkian hunters. 
Here, in addition to the difficulties engendered by the fundamental 
ambiguity of current psychiatric theory over the respective causal 
roles of psychological and organic factors in clinically familiar syn- 
dromes, there are (or ought to be) serious uncertainties introduced 
by recognition of the extreme climatic, epidemiological (in respect 
to infectious diseases) , and nutritional conditions to which tech- 
nologically primitive populations are at times exposed (see, for 
example. Tooth's discussion of the difficulty even psychiatrists ex- 
perience, when using purely behavioral criteria, in making the 
differential diagnosis between schizophrenia and certain types of 
trypanosomiasis in West Africa) (Tooth, 1950) . 

This paper is not intended, however, as an admonition to anthro- 


pologists to abandon an obsolete dogma for the sake of embracing 
a new scientific faith. Rather, the necessity for incorporating a new 
viewpoint into an existing tradition is pointed out. That this incor- 
poration will entail modification of some beliefs and procedures 
may be expected; but the new theoretical position should be a strong 
synthesis rather than a weak substitute. 


The year 1927 may be taken as the beginning of codification of 
the culture and personality position in anthropology, for in that 
year Sapir's pioneer paper, "The Unconscious Patterning of Be- 
havior in Society," was published in a symposium on The Uncon- 
scious (Mandelbaum 1949). Sapir's paper, probably the first 
major piece of theoretical writing in the culture and personality 
tradition, set, or at least prefigured, the frame of reference of later 
anthropological work in this area. This frame of reference was pre- 
dominantly psychological rather than biological: it implied that 
the fundamental, and often unconscious, organizations of indi- 
vidual behavior which are conventionally labeled "personality" are 
molded, not by physical constitution, but by a combination of 
cultural milieu and individual experience. The correspondingly 
functional character of the conventional culture and personality 
view of mental disorder, as it developed in the next few years in the 
work of Sapir, Benedict, Mead, and others, can be readily explained 
by the absence of any substantial competing body of thought; for 
the biological approach in psychiatry did not even begin to make 
headway until after 1927. 

The most impressive body of psychiatric theory in 1927 was 
psychoanalytic. This theory, although it gave lip service to biologi- 
cal thinking, and although its builders were well grounded in 
neurology, was in operation uncompromisingly psychological. Ac- 
cordingly, the published case histories provided very little informa- 
tion concerning the physiological status of the patients. The analyst 
sometimes used physical metaphors (like "the economy of psychic 
energy"), invoked constitutional predispositions, and made as- 
sumptions about organically grounded instincts, erogenous body 
zones, and stages of sexual maturation. Freud, himself a neurologist 
of distinction, even asserted that behind the analyst stood the man 
with the syringe. But the psychoanalytic physiology, as it grew 
beyond Freud's control, was increasingly a pseudophysiology. 
Biological man was for all practical purposes constant in the psy- 


choanalytic equation, and "psychological" events (learnings, com- 
munications, fantasies, motives, defense mechanisms, etc.) were the 

Most of the currently prominent "organic" methods of treatment 
were developed after psychoanalysis reached its theoretical matu- 
rity. In 1927 psychiatry had little else to oflfer in treatment beyond 
psychological (including psychoanalytic) methods for the well- 
to-do and custodial care (eked out by sedatives, hydrotherapy, and 
work therapy) for the poor. The insulin coma treatment for schizo- 
phrenia was introduced about 1930 and metrazol convulsive ther- 
apy in 1936; electroshock was not developed until 1938 (and all of 
these treatments were first publicly described in Europe) . Psycho- 
surgery was seriously developed in Portugal about 1935 and in this 
country in 1936. Psychopharmacology, hitherto a somewhat exotic 
specialty, began to flourish only during World War II. The use of 
drugs for abreaction of emotional conflict in combat neuroses be- 
came prominent during the early years of the war; and the intensive 
study of the psychotomimetic drugs (principally hallucinogens) 
and their experimental use for therapeutic purposes has developed 
chiefly since World War II. The new tranquilizing (or "ataractic") 
drugs were first offered to the medical profession in 1952, and the en- 
ergizers (or "psychostimulants") have come even later. 

Basic science contributions, apart from psychoanalytic theory, 
were equally uninspiring in 1927. Inspired by the discovery of the 
role of syphilis in paretic psychoses, early speculations about the role 
of focal infection in the etiology of the other psychoses were failing 
to find clinical confirmation. Berger's first report on the use of the 
electroencephalograph (EEG) for recording "brain waves" (elec- 
trical potentials originating in the cerebral cortex and in other parts 
of the brain as well) was not published in Germany until 1929; not 
until 1935 did American scientists publish confirmatory findings. 
Clinical chemistry had only in the preceding fifteen years developed 
the basic techniques for analysis of small samples of blood; prior to 
World War I, investigations of human metabolic processes had had 
to depend largely on studies of diet and urine, because the quantities 
of blood required for chemical analysis were so large as to prohibit 
their use as routine clinical procedures. The application of these new 
techniques of blood analysis to problems of psychiatric research, 
and the biochemical findings based on their use, came almost en- 
tirely after 1927. Thus, for instance, endocrinology was still in its 
infancy in 1927. The importance of the hormones of the adrenal 


cortex, which play a role in regulating the carbohydrate metabolism 
and the balance of mineral electrolytes in the body fluids, and which 
in excess can precipitate psychotic states, was not realized until the 
late 1920's. Research in that area was so slow in diffusing into other 
branches of knowledge that as late as 1944, in a widely read two- 
volume symposium entitled Personality and the Behavior Disorders 
(Hunt 1944), the adrenal cortex is given one paragraph (and no 
mention in the index.) Thus Selye's first publication on the cele- 
brated stress or general adaptation syndrome concept was first pub- 
lished in Nature in 1936 {vide Selye 1956); and the "cortisone 
psychoses" did not even exist until cortisone was isolated, synthe- 
sized, and finally used in the treatment of arthritis about 1945. 
Franz Kallman's early report on his genetic studies of schizophrenia 
utilizing pairs of identical twins was published in 1938 (Kallman 
1938). The more modern theories of nerve impulse transmission 
emerged during and after World War II, some of them stimulated 
by investigations into the action of the so-called "nerve gases" by 
the Army Chemical Center. 

But there is no reason to continue the demonstration farther. The 
major point is clear: a large part of the modern knowledge of the 
physiological parameters of the behavior of the central nervous sys- 
tem in man has been accumulated since the original conceptual 
structure of the culture and personality viewpoint was built by 
Sapir, Mead, and other pioneer scholars. Whole literatures, rivaling 
in size the entire body of culture and personality writings, now exist 
on such topics as the relation between the adrenal hormones and 
mental function, the localization of labor in the brain as revealed by 
electroencephalographic and derivative techniques, and the effects 
of drugs on mood and cognitive process. And the major portion of 
all of these fields of knowledge has been contributed well after cul- 
ture and personality committed itself to a functional approach. 

As yet, the various special lines of the new organic approach have 
not achieved synthesis either among themselves or with the (actually 
older) psychosocial tradition in psychiatry and the social sciences. 
Nevertheless, a general philosophy would seem to animate the ap- 
proach and to determine the nature of any future synthesis with 
the functional position. This philosophy would seem to reside in four 

I. Statements about "behavior," "mind," "personality," "psyche," "mental 
illness," and other "psychological" entities are statements about physical systems 
which include brain (for the brain is the mind). 


2. Any physical disfunction of brain implies some mental disfunction. 

3. Some physical disfunctions will produce disorganizations of neural systems 
most of whose components will remain individually undamaged. 

4. Most cases of chronic, and many of acute, behavior disorders (including the 
functional psychoses) are the symptomatic consequences of chronic, or acute, 
physical disfunctions of brain. 

The reader will note that the organic approach, as thus stated, does 
not claim that every socially undesirable mental state, attitude, or 
motive necessarily implies a physical disfunction; thus, evidences of 
hostility and anxiety, " neurotic" defenses, suicide, antisocial acting 
out, and so forth may in principle be produced by brains which 
function perfectly well but have been subjected to environmental 
pressures (including faulty communication) to which these "symp- 
toms" are "normal" responses. But the organic approach would dif- 
fer from the functional approach in claiming that an adequately 
functioning brain will be able to adapt to, or reduce, environmental 
pressures, and that chronic mental disfunctions are therefore pre- 
ponderantly the consequence of a chronic physical disfunction 
which existed prior to, or independently of, the organism's embar- 
rassment by environmental pressures. A radical functional theory, 
by contrast, would ascribe a far smaller role to organic factors as 
causal agents in all except the gross and obvious types of organic 
brain damage; but most functionalists would probably concede that 
chronic psychogenic stress can on occasion elicit physiological 
alterations, sometimes irreversible, which aggravate functional 
mental disorders (just as chronic psychogenic stress can lead to non- 
mental organic disorders such as duodenal ulcer) . 

More specifically the organic approach can be divided into such 
main topical areas as: 

1. The study of the anatomy and physiology of the central nervous system 
(including the autonomic system) considered as an entity. 

2. The study of the localization and organization of labor in brain (including 
the logical structure of nerve nets) . 

3 . The study of nerve and nerve impulse. 

4. The study of the relation of metabolic (including digestive, excretory, 
circulatory, endocrine, and intracellular biochemical) processes to cerebral func- 

5. The study of the genetics of mental disorders. 

6. The study of the effect of hypoxia, hypoglycemia, and electrolyte imbalance 
on cerebral function and the various processes responsible for hypoxia, hypogly- 
cemia, and electrolyte imbalance. 

7. Psychopharmacology (including the study of tranquilizers, energizers, and 
psychotomimetic agents). 


8. The study of the eflfect of nutritional variables on cerebral function. 

9. The study of the shock therapies (principally insuUn coma and electro- 
shock) . 

10. The search for blood fractions containing suspected psychopathogenic 
(toxic) substances spontaneously produced by the body. 

The disciplines involved in these and other studies of psychopa- 
thology range from mathematical physics and computer design, 
through such laboratory sciences as physical chemistry, biochemis- 
try, clinical chemistry, physiology, experimental psychology, and 
neuropsychiatry, to those areas of anthropology and sociology 
which can contribute data, method, or theory to organically 
oriented investigations. 

A major problem in the organic approach has, of course, been its 
relative insularity from psychosocial knowledge (this has not been 
a problem of the functional approach alone) . Accordingly a major 
need of both approaches is a better understanding of how knowl- 
edge and speculation concerning the physical aspects of human 
systems can best be related to knowledge and speculation concern- 
ing the psychological and social aspects of these systems. This is 
imperative because, although cases of mental illness are usually first 
identified in the community by laymen using social criteria rather 
than criteria of physical science, and although some part of the 
total disease process is invariably a function of social system inter- 
acting with individual personality, if the development of many of 
these cases is dependent on organic processes, then very careful 
analysis must be made of the interaction of social and organic 
events. And anthropology, by both theory and field investigation, 
can contribute significantly to the advancement of this kind of 


In its simplest form, the problem faced by anthropological the- 
ory in the area of mental illness can be illustrated by the syndrome 
pibloktoq among the Polar Eskimo of the Thule District of north- 

^ The description of the pibloktoq syndrome is based on a compilation of published and manu- 
script descriptions, both specific and generalized, by a variety of observers, from the missionary 
Hans Egede in 1765 to about 1940. Seventeen photographs of a woman during a pibloktoq attack 
at Etah were taken by Donald MacMillan in June 19 14; we were able to use copies of these 
from the original negatives on file in the Photographic Division of the American Museum of 
Natural History. I am indebted to Mr. Robert Ackerman, my collaborator in the pibloktoq 
study, who has collected many of the data and contributed heavily to their interpretation; to 
Dr. Zachary Gussow, who kindly permitted use of his unpublished manuscript on pibloktoq; 
and to Dr. Gilbert Ling, who reviewed the calcium hypothesis and contributed to its refinement. 


ern Greenland. The classic course of the syndrome, as judged from 
cases described by various travelers in the north (MacMillan 1934; 
Peary 1907; Rasmussen 191 5; Whitney 191 1) and from photo- 
graphs of one attack (American Museum of Natural History 
19 14), is as follows: 

1 . Prodrome. In some cases a period of hours or days is reported during which 
the victim seems to be mildly irritable or withdrawn. 

2. Excitement. Suddenly, with little or no warning, the victim becomes 
wildly excited. He may tear off his clothing, break furniture, shout obscenely, 
throw objects, eat feces, or perform other irrational acts. Usually he finally leaves 
shelter and runs frantically onto tundra or ice pack, plunges into snowdrifts, 
climbs onto icebergs, and may actually place himself in considerable danger, from 
which pursuing persons usually rescue him, however. Excitement may persist for 
a few minutes up to about half an hour. 

3. Convulsions and Stupor. The excitement is succeeded by convulsive sei- 
zures in at least some cases, by collapse, and finally by stuporous sleep or coma 
lasting for up to twelve hours. 

4. Recovery. Following an attack, the victim behaves perfectly normally; 
there is amnesia for the experience. Some victims have repeated attacks; others 
are not known to have had more than one. 

The epidemiological parameters seem to be: 

1. Geographical. Pibloktoq (or, in Danish usage, perdlerorpoq) is known to 
occur among the Polar Eskimo of the Thule District. Whether the same syndrome 
(whatever it is called) occurs elsewhere is uncertain. Hoygaard, in a dietary and 
medical study of the Angmagssalik Eskimo in 1936-37, reported that "^^ Hysterical 
fits accompanied by strong mental and physical excitation were frequent, espe- 
cially in women" (Hoygaard 1941:72) . It does not seem to have been noted, how- 
ever, among Canadian or Alaskan Eskimo, nor is it certain that it occurs in Asia 
or northern Europe. Thus we can only say that it certainly occurs in northwest 
Greenland; that it probably occurs elsewhere in Greenland; and that it may occur 
anywhere in the world. Whether or not the syndrome is to be considered a uniquely 
arctic or even Polar Eskimo affliction depends on whether it is a unique disease. 

2. Seasonal. Reports describe cases occuring at all seasons of the year but 
cases are said to be fewer in the summer. 

3. Historical. As might be expected, since the Thule Eskimo were not visited 
by white men until 18 18, the case notes and descriptions are recent, the best of 
them dating from the time of Peary's visits to the Polar Eskimo in the first decade 
of the twentieth century. Detailed accounts have been provided by Peary ( 1907) , 
MacMillan (1934), Knud and Niels Rasmussen ( 1 9 1 5 ) , and Gussow (i960), and 
others familiar with the Polar Eskimo. It is probable, however, that the disorder 
is fairly ancient in the area. As early as the mid-eighteenth century, northwest 
Greenlanders (possibly including the Polar Eskimo) were reported to be peculiarly 
subject to the "falling sickness." And in the 1850's the crew of Kane's icebound 
ship, twice wintering north of Thule, were afflicted by a strange "epilepto- 
tetanoidal disease" which, in combination with scurvy, killed at least two men. 


incapacitated others, and rendered their dogs worthless (Kane, 1856). "Epilepto- 
tetanoidal" is a reasonably accurate descriptive phrase for pibloktoq. 

4. Frequency. Pibloktoq can apparently reach epidemic proportions: eight of 
seventeen Eskimo women associated with Peary's 1908 expedition were afflicted 
during one winter season; other observers have claimed that at certain times cases 
could be seen almost every day in a single village. 

5. Racial Nonspecificity. As was noted above, several probable cases of 
pibloktoq among scorbutic whites were observed by Kane and Hayes in the 1850's 
in the same region. 

6. Possible Species Nonspecificity. "Fits" among sled dogs, with social with- 
drawal, snarling, fighting, and convulsive seizures, but usually ending in death, 
are said to be regarded by Eskimo as the same syndrome and are given the same 
name, pibloktoq, as the human attacks. 

The Hysteria Hypothesis 

The major psychological explanation of the pibloktoq syndrome 
has been psychoanalytic. In 191 3 A. A. Brill, Freud's self-appointed 
American apostle, wrote a paper on the subject based on a reading 
of one of Peary's books and on personal discussion with Donald 
MacMillan, the naval officer who accompanied Peary (Brill 1913) . 
Brill considered the syndrome to be classic hysteria major. Following 
a somewhat simplified Freudian model, he interpreted the seizures 
as expressions of frustration at lack of love and cited as the type 
case a female who displayed particularly flamboyant attacks. This 
attractive young woman had not succeeded in getting a husband 
because she was a poor seamstress; she was consequently frustrated 
in her emotional need for love in all but the most crudely physical 
sense. More recently, Gussow (i960) has extended Brill's formula- 
tion, interpreting the hysterical flight as a seductive maneuver, an 
"invitation to be pursued," in persons whose chronic insecurities 
have been mobilized by some precipitating loss or fear of loss, and 
who seek loving reassurance in a "primitive and infantile, but char- 
acteristically Eskimo, manner." Indeed, he feels that such reactions 
are a manifestation of the basic Eskimo personality. The greater 
frequency of pibloktoq in women he explains culturally as the 
result of "the socially subservient position of women . . . and their 
added helplessness in the face of culturally traumatic experiences." 
The nudity is in part explained by the common tendency of Eskimo 
to undress indoors and to chill the naked body out of doors after 
the sweat bath. The glossolalia, mimetic behavior, shouting, weep- 
ing, and singing sometimes observed he also explains culturally by 
pointing out that these behaviors are found in shamanistic per- 
formances and religious ceremonies, not only among the Eskimo, 
but also in Korea. The flight is considered to be a hysterically moti- 


vated invitation to be taken care of, rather than a component of 
an involuntary psychomotor seizure pattern, because no cases of 
flight have been reported in which the victim was not seen, fol- 
lowed, and rescued. The asserted tendency for pibloktoq to occur 
in winter is illuminated by the observation "that winter, more than 
other seasons, intensifies Eskimo insecurity — and hence their prone - 
ness to derangement — through increased threat of starvation, high 
rate of accidents, fear of the future, and so forth." 

These psychoanalytic and psychocultural explanations, however, 
are for several reasons not entirely satisfying. Nudity, for instance, 
is indeed culturally prefigured, since it is the only means of reducing 
body temperature in persons who have no clothes to wear other than 
heavy furs in poorly ventilated dwellings where the temperature 
may rise to over ioo° F. But this suggests that the denudation may 
be merely a response to a sudden somatic sensation of extreme heat. 
The fact that most reported victims of hysterical flight were res- 
cued from danger without injury may obviously be an artifact of 
observation: any victims who froze, drowned, lost themselves, were 
carried away on drifting ice, fell and died alone in the snow, and so 
on, would by definition be those who were not observed. Further- 
more, in at least one case, a rescued woman tvas injured; she suf- 
fered a frozen hand and breast, a serious condition in the absence of 
European medical technology. Two of Kane's men died and the 
dogs often die. Glossolalia, singing, and so forth are hardly evidence 
for an influence of Eskimo culture on the form of this hysteria, 
since these behaviors are virtually pandemic. The evidences of ex- 
treme physiological stress (bloodshot eyes, flushing of face, foam- 
ing at mouth, convulsive movements) and the demented behavior 
(attempting to walk on the ceiling, eating of feces, and ineffectual 
destructiveness) are not prefigured in the culture. And finally, the 
Eskimo are not reported to explain these fits (in contrast to psy- 
chotic disorders) by supernatural theories of disease (such as pos- 
session, witchcraft, punishment for taboo violation, or soul loss) 
but seem to regard them as natural ailments, experienced by dogs 
and men alike, comparable perhaps to the common cold, the broken 
limb, and other ills that the flesh is heir to. This phlegmatic response 
would not provide very much in the way of reward for a hysterical 

The Calcium Deficiency Hypothesis 

An alternative, and in part biological, hypothesis can be sug- 
gested which explains pibloktoq with at least equal plausibility. 


Low concentrations of ionized calcium in the blood (hypocal- 
cemia) produce a neuromuscular syndrome known as tetany which 
is often complicated by emotional and cognitive disorganization. 
The neurological symptoms of tetany include characteristic mus- 
cular spasms of hands, feet, throat, face, and other musculature, 
and in severe attacks, major convulsive seizures. The tetanic syn- 
drome may be precipitated by trivial stimuli and is usually brief 
and sporadic rather than continuous (continuous tetany may of 
course be fatal) . Although the information available in the photo- 
graphs and literature is not sufficient in itself to establish the diag- 
nosis, the symptoms of pibloktoq are compatible with the clinical 
picture of hypocalcemic tetany, and several authorities have sug- 
gested the calcium deficiency hypothesis (Hoygaard 1941:72; 
Baashuus- Jensen 1935:344, 388; and Alexander Leighton in a per- 
sonal communication) . Observation and testing in the field would 
be required to confirm the hypocalcemic hypothesis and to rule out 
alternative diagnoses (hypoglycemic shock, hysteria, food poison- 
ing, virus, encephalitis, etc.). It is also possible that a tendency 
toward epilepsy may have been genetically determined by inbreed- 
ing in this small isolated group; this is suggested by reports that 
epilepsy is more common in northern Greenland than elsewhere on 
the island. The hypocalcemia and epilepsy theories are not mutually 
exclusive, however, since hypocalcemia probably would tend to 
precipitate a latent seizure in persons prone to epilepsy. Observation 
and testing for differential diagnosis would require both the elicit- 
ing of neurological signs in victims during attack, or in persons with 
a history of attacks, and blood tests on victims and on samples of 
pibloktoq-prone and piblokfoq-iree persons for serum calcium, 
serum potassium, and possibly other constituents. 

The plausibility of the calcium deficiency hypothesis is supported 
not merely by the opinions of certain authorities and by the com- 
patibility of the pibloktoq syndrome with the syndrome of hypo- 
calcemic tetany, however. It is also suggested by indirect evidence, 
both medical and ecological. 

Medically, the Eskimo of Greenland (including the Thule Dis- 
trict) are characterized by a proneness to hemorrhage and slow 
coagulation (Hoygaard 1941:83-85, and Cook 1894:172). Such a 
tendency toward bleeding might conceivably be associated with low 
serum calcium levels (although vitamin K deficiency is more likely 
to lead to this condition) . At Angmagssalik, convulsions in infants, 
suggestive of hypocalcemic tetany, were reported by Hoygaard to 


be frequent (Hoygaard 1941:78, 135), and Bertelsen noted in a 
medical report on the Greenland Eskimo that there was a high 
frequency of cramps, especially of the legs, even in adults (Bertel- 
sen 1940:216) . These observations are reminiscent of the account 
by Kane of the "strange epilepto-tetanoidal disease" which inca- 
pacitated his crew north of Smith Sound in the 185 o's. He diagnosed 
two fatal cases of "tetanus" displaying laryngospasm (these could 
have been actually hypocalcemic tetany going into statics eclainp- 
ticus) , two fatal cases of the "epilepto-tetanoidal disease," and 
numerous cases of cramps and muscular pains, sometimes accom- 
panied by "mental symptoms" of disorientation and confusion, 
both in dogs and man (Kane 1856). 

Ecologically, it may without hesitation be stated that the high 
arctic environment does not provide rich sources of nutritionally 
available calcium during all seasons of the year to technologically 
primitive populations. Hoygaard found that nearly half of the an- 
nual calcium intake at Angmagssalik was provided by dried capelin 
(the bones of dried capelin being edible) . When dried capelin was 
available, the calcium intake was low but above the level asserted 
by medical authorities to be the minimum for maintenance of 
health. But without dried capelin (a circumstance which periodi- 
cally occurred as a result of unavailability of the fish or unsuitability 
of the weather for drying them) , calcium intake dropped well be- 
low the minimum (Hoygaard 1941 ) . Rodahl also found the dietary 
of certain Alaskan Eskimo groups to be relatively low in calcium 
(Rodahl 1957) . At Thule, although no careful dietary studies have 
been found, it is reported that little fishing is done because fish are 
sparse and consequently capelin is not caught in substantial quan- 
tity. Probably substituting for dried capelin, however, are birds — 
the "little auks" — which, after storage in seal oil, can be eaten whole, 
including, apparently, some of the bones (MacMillan 1918). A 
further ecological complication may be a product of the high lati- 
tude itself. Man requires a certain quantity of vitamin Da in order 
to absorb and utilize dietary calcium efficiently (and possibly also 
to metabolize carbohydrate efficiently) . This vitamin is formed in 
the human and animal skin when ultraviolet light activates cer- 
tain cholesterol-containing oils. In the high arctic, however, a com- 
bination of low sun angle during summer, a long period of winter 
darkness, and the need for heavy clothing during most of the year, 
must prevent the human body from synthesizing much of its own 
vitamin Da. Whether sufficient vitamin Ds can be secured from sea 


fauna at this latitude is uncertain. Seal oil contains significant quan- 
tities of vitamin Da but, at Thule, the fish oils rich in vitamin Da, 
such as cod liver oil, are probably not a major source of supply be- 
cause of the aforementioned lightness of fishing in that region. To 
summarize the ecological problem briefly, even if sufficient vitamin 
Ds is available to allow maximum efficiency in calcium absorption 
and utilization, it is still highly probable that some people, at some 
seasons of the year, will be unable to secure sufficient dietary calcium 
to meet published medical standards. If such a low calcium intake 
were coupled with a high protein and high potassium intake, the 
neurological consequences would be intensified, and the heavy meat 
consumption of Polar Eskimo entails a large intake of protein and 

One fact, however, militates against a simple dietary calcium 
deficiency hypothesis: the reported extreme rarity of rickets in 
Eskimo infants and of osteomalacia in Eskimo adults (for example, 
in pregnant and lactating women) (Bertelsen 1940). These are 
diseases in which, as a consequence of inadequate calcium intake 
or utilization, or both, the bones yield their calcium to the blood 
and, eventually, to the urine, with the sufferer thus gradually losing 
calcium from the body at the expense of bony tissue. In temperate 
latitudes, rickets and osteomalacia are normally forestalled by milk, 
sunlight, and supplementary vitamin Ds preparations in cod liver 
oil and vitamin pills. If one hypothesizes that the Eskimo diet is 
low in calcium, and perhaps in sun-formed vitamin Ds, how is it 
that rickets is not evident? The answer to this question requires 
another hypothesis concerning hormonal function. It would seem 
that if calcium and/or vitamin Ds intake is chronically low in the 
high arctic environment, then the Eskimo physiology must for 
generations have been forced to "choose" between tetany and 
rickets — and, unlike more southerly populations, it has "chosen" 
tetany as the lesser of two evils. (More precisely, of course, it is 
the environment which has selected the better-fitted physiological 
alternative. ) Rickets and osteomalacia would in a primitive Eskimo 
economy be fatal because they are physically crippling. Sporadic 
attacks of tetany, even if occasionally damaging or even fatal, 
would be by comparison merely an annoyance. Hence the hypo- 
calcemia hypothesis requires the corollary that the Polar and per- 
haps other Eskimo tend to be mildly hypoparathyroid (or, more 
exactly, again, that in this cultural-ecological matrix, optimum 


parathyroid function requires a lower activity than does optimum 
function under the conditions of European and American medical 
practice) . Such a mild "hypoparathyroidism" would be conceived 
as a product of natural selection for primitive life in an arctic en- 
vironment, yielding a type of hormonal balance which retains cal- 
cium in the bones even if calcium levels in serum fall occasionally. 
There is, as a matter of fact, some evidence to support this hypothe- 
sis. The doomed medieval Norsemen, not preadapted to a high arctic 
environment, who settled along the west coast of Greenland, and 
who finally died out and were replaced by ricketless Eskimo, did 
suffer from rickets and osteomalacia (Maxwell 1930:20). 

But if we propose a hypocalcemia hypothesis, do we ignore Es- 
kimo culture? Certainly not. Consideration of cultural factors is, 
in fact, already implicit in the hypothesis as enunciated. This hy- 
pothesis rests on the assumption that the subsistence technology is 
"primitive," that is, in this application of the concept, that manu- 
factured vitamins and imported or specially processed calcium- 
containing foods are not available and that, to hunters, a strong and 
undistorted skeletal structure is of greater survival value than free- 
dom from occasional attacks of tetany. These cultural characteris- 
tics render the population vulnerable to a local dietary calcium 
and/or vitamin Ds shortage and select the nervous and muscular 
system rather than the skeleton as the target tissue of any calcium 
and/or vitamin Ds nutritional deficiency. 

But Eskimo culture also functions to minimize, within the limits 
stated above, the frequency and severity of attacks, via the customs 
of securing, processing, and storing of large quantities of calcium- 
containing birds (the "little auks") ; of obtaining, preserving, and 
making extensive use of vitamin-Ds-containing seal oils; of strip- 
ping and exposing the body to direct sunlight whenever the weather 
permits; of weaning children late (thus ensuring them maximal cal- 
cium intake in mother's milk during the rickets-vulnerable period 
of infancy) ; of securing to pregnant women (who are particularly 
vulnerable to osteomalacia) and children preferred access to fresh 
and stored foods high in calcium (specifically, the little auks and 
whatever dried fish are available) by making women and children 
chiefly responsible for netting the birds and collecting the eggs, and 
(to judge from taboos reported from Eskimo groups other than 
Thule) by maintaining food taboos which have the effect at cer- 
tain times of substantially restricting the pregnant or lactating 


mother to the use of dried fish, birds, or other stored foods high 
in calcium. 

It is possible that, apart from its role in etiology, Eskimo custom 
also affects the details of overt symptomatology. Conceivably the 
frequently reported impetuous flight from the group during the 
initial phases of an attack may reflect a personality trait common 
among Eskimo: withdrawal from, rather than aggression in, a situ- 
ation when the individual's confidence in his ability to master it 
has been shaken. Such a tendency may be reflected in the tendency 
for Eskimo men to abandon kayak hunting if their confidence has 
once been disturbed ("kayak-phobia") ; by the practice of kivik- 
toq, or "going into the mountains" to live a hermit's life, in men 
and women alike who feel rejected by their communities; by the 
reported willingness of the aged and infirm to be abandoned to die; 
and by the anxiousness of Eskimo parents not to disturb the con- 
fidence of their children, even when playing dangerously, by frus- 
trating negative commands. Such a psychological interpretation — 
which is, in a sense, directly contradictory to the hysteria hypothesis 
— rests on the assumption that any incipient neurological disfunc- 
tion is susceptible to different interpretations by the victim and his 
associates and can therefore precipitate different overt responses, 
depending on particular customs of the individual and group. 

And finally, with regard to its handling of cases of pibloktoq, 
Eskimo custom obviously plays a very important role. An attack of 
pibloktoq is not automatically taken as a sign of the individual's 
general incompetency. The victim is, if necessary, prevented from 
injuring himself or others; otherwise he is left alone while the at- 
tack spends itself. The attack may be the subject of good-humored 
joking later but is not used to justify restriction of the victim's so- 
cial participation. There is, in other words, little or no stigma; the 
attack is treated as an isolated event rather than as a symptom of 
deeper illness. Such a phlegmatic approach would seem well calcu- 
lated once again to minimize any damage to the individual's per- 
sonal confidence and thus would work to forestall the development 
of chronic psychological invalidism. The impact on chronicity of 
differential handling of such episodic disorders is well illustrated in 
the history of American combat psychiatry, which between World 
War II and the Korean War achieved a 50 per cent reduction in the 
rate of chronic psychoneurosis developing out of combat break- 
down simply by refusing to treat the breakdown as a symptom of 
illness (Glass 1953)- 


Implicotions of the Alternotive Theories 

Two alternative armchair theories of pibloktoq have been pre- 
sented. Although the "organic" (hypocalcemia) theory seems 
preferable, the organic theory is just as much concerned with analy- 
sis of cultural factors as is the "psychological" (hysteria) theory. 
In order to choose between the two, field investigation will be 
necessary. Such field investigation will have considerable signifi- 
cance for anthropological theories of mental illness (and profes- 
sional psychiatric theory, for that matter) . For not only will it 
contribute to the solution of a particular — and to some eyes, per- 
haps, an unnecessarily exotic diagnostic problem, it will also bear 
on two major theoretical issues. 

One of these major issues is the understanding of hysteria itself. 
As is well known, psychoanalysis was originally conceived as a 
means for treating hysteria, and upon the analysis of cases diag- 
nosed as hysteria much of its theoretical structure has been erected. 
Since Freud's time, hysteria has become a rare disorder in most of 
Europe and America. This may be the consequence of culturally 
determined changes in modal personality structure in Western 
countries and in preferences for various styles of psychosomatic 
expression. It may also be the result of changes in diagnostic prac- 
tice (it has been suggested, for instance, that "hysteria has van- 
ished right into the diagnosis of epilepsy" (Peterson 1950) ) . And 
it may be the result of culturally determined changes in such mat- 
ters as style of dress and housing, hours of work, methods of light- 
ing, and diet, which could affect, in particular, calcium intake and 
utilization in persons vulnerable to tetany and rickets. Certainly 
rickets has become more rare in precisely those groups once most 
prone to grand hysteria: the Western European urban populations. 
But now we are suggesting that at least one type of hysteria (the 
"grand hysterical attack") may not be purely psychogenic! 

Such an implication demands support by way of empirical in- 
vestigation — an investigation which, in fact, takes up again an 
abortive line of inquiry into the relationship between tetany and 
hysteria that began in Europe before the psychoanalytic theories 
of hysteria swept competing approaches from the field (Barrett 
1919—1920:385-386). It is of more than antiquarian interest to 
recall that between 1880 and 1895 there was a veritable endemic of 
tetany among the working class of Vienna, Paris, and other Euro- 
pean cities (Shelling 1935:115—116). This plague of tetany was. 


at the time, not understood etiologically, for the role of calcium in 
tetany had not been established. During the same period, the work 
of French and Viennese neuropsychiatrists on hysteria was being 
pursued most intensively, and it culminated, as everyone knows, in 
Freud and Breuer's Studies in Hysteria, which was published in 
1895 after a preliminary publication in 1893. This study revealed 
the psychological connection between the hysterical symptom and 
traumatic emotional conflict and suggested a technique of "talk- 
ing" therapy which soon developed into the method of psycho- 
analysis. We might now ask, however, whether the physiological 
milieu of hypocalcemia may not have been a conditioning factor in 
hysteria. The most serious endemics of rickets and of hypocalcemic 
tetany — determined by constraints of custom and/or economy on 
food, dress, interior lighting, working hours, and access to open 
spaces not only among working people but among all classes in late 
nineteenth century Europe — came at precisely the same time that 
hysteria reached its peak as a psychiatric problem. The discovery 
of the value of sunlight, milk, and vitamin-Ds-containing foods, 
and the general amelioration of social conditions, during the early 
twentieth century, was accompanied by a drastic reduction in the 
frequencies of rickets, of tetany, and of hysteria. Thus we may 
suggest, as a hypothesis for medicohistorical investigation, that the 
hysterical attack and perhaps even hysterical conversion will occur 
most readily in persons with low levels of serum ionized calcium and 
that chronically low levels may maintain a neurophysiological 
milieu in which either tetany, hysterical attacks, hypersuggesti- 
bility, or hysterical learning of conversion symptoms is sooner or 
later inevitable, the choice of disorder depending on various con- 
ditioning factors of situation, personal history, and biochemical in- 

Suggesting that the late nineteenth century European hysterias 
may have been in considerable proportion undiagnosed cases of 
serum calcium deficiency raises a major issue in psychiatric theory, 
for psychoanalysis was founded on the analysis of hysterics. In view 
of this fact, it may be well to evaluate further the culture-historical 
dimensions of the issue. The late nineteenth century students of 
hysteria — including Freud — were aware that hysterics might dis- 
play unusual physiological profiles as well as disordered behavior, 
and some felt that hereditary predisposition played a role in the 
pathogenesis of the disease. But these psychiatrists of the 1890's 
were in somewhat the same position vis a vis physiological explana- 


tions of hysteria as the anthropologists of the 1920's were vis a vis 
explanations of psychopathology in general: physiological investi- 
gations had not advanced far enough to provide a base for framing 
testable physiological hypotheses. 

Thus the first demonstration that tetany was associated with re- 
duced concentration of calcium in the blood was not made until 
1908; hitherto the diagnosis depended on the finding of positive 
neurological signs. Not until 1921 did the development of micro- 
metric methods of determining quantities of serum calcium make 
possible widespread testing for serum calcium level (Shelling 1935: 
114— 116). Differential diagnosis in certain cases between hysteria 
and tetany was extremely difficult, and in fact probably was arbi- 
trary, before the development of the serum calcium and tetany 
hypothesis and the provision of appropriate methods of clinical 
chemistry. Consequently, some cases which today would probably 
be regarded as unequivocally tetany (e.g., the tetanic syndrome 
following thyroidectomy) were in 1904 diagnosed as mixtures of 
tetany and hysteria (cf. Curschmann 1904) . Thus it is impossible 
that Freud could have considered the possibility that hysteria might 
be a symptomatic consequence of low serum calcium. The cultural 
milieu in which he worked had not provided him with the concepts 
or tools by which the question could have been asked or answered. 
Inasmuch as we cannot return to the nineteenth century to do 
serum calcium determinations on Freud's original patients, we can- 
not say what the results would have been, nor can we estimate the 
impact on the development of psychoanalysis if the findings had 
been positive. But at least we have still another historical answer to 
the question "Why has hysteria virtually disappeared in Europe and 
the United States?" Our (metaphorical) answer is, "It dissolved 
in bottles of milk and cod-liver oil" — that is to say, the cultural 
changes associated with an appreciation of the importance of sun- 
light, vitamin Ds, milk, and various other factors for maintaining 
proper calcium balance, together with a general improvement of 
nutritional standards, has virtually eliminated (except in certain 
rare medical conditions) a total syndrome, one symptom cluster of 
which was once (and still is) called tetany, and another symptom 
cluster of which was once (but no longer is) called "grand hysteri- 
cal attack." 

The need for empirical evidence bearing on the hypotheses out- 
lined above leads immediately to a consideration of the second major 
issue: the larger theoretical structure which should guide such an 


investigation. It is evident that even if it is possible to identify a 
specific physiological variable as the precipitant of the overt 
symptomatology, an adequate explanation of the frequency of the 
syndrome in the population, its geographical range, its racial and 
species distribution, its seasonal variation, its history, and the sever- 
ity and details of form of the symptoms themselves, must depend 
on evaluating other variables, physiological, psychological, and 
cultural. It is the interaction of these other variables with the im- 
mediately precipitating physiological variable which provides the 
necessary and sufficient conditions for a type of mental illness to 
occur in a particular group with a particular frequency. We have 
already suggested some of these conditions in the pihloktoq analy- 
sis. Let us now turn our attention to the development of a frame 
of reference which can guide the refinement of theory and the 
acquisition of relevant empirical data. We shall begin, in the next 
section, with a further discussion of a point introduced in the 
pibloktoq analysis: the importance of the "theory of illness" in the 
formation of a symptomatic structure. And finally we shall at- 
tempt to generalize the line of thought represented in the pibloktoq 
analysis, and in the following discussion, into a rough model of a 
biocultural approach to mental illness. 




Mental illness is an episode in a life program, usually following a 
more or less extended period of normalcy (as defined by both the 
person and his community) , and terminated either by death or by 
a return (temporary perhaps) to normalcy. In the biocultural 
model, a conjunction of pathogenic, organic, and psychological 
events is considered to abort a life program normal to the society 
by crippling the victim's apparatus for cognitive organization. 
With the onset of the physiologically determined desemantication 
(reduced cognitive organization capacity) the victim is unable to 
organize his perceptions, his motives, and his actions meaningfully 
so as to satisfy his own wishes without frustrating those of others 
or vice versa. His more or less desperate efforts to protect himself 
from the consequences which he expects to follow the drastic re- 
duction of cognitive capacity are apt to be the most conspicuous 
symptoms of the disorder: withdrawal, aggression, paranoid delu- 
sion, and the bizarre use of the familiar mechanisms of defense like 


repression, sublimation, denial, etc. And simultaneously, the vic- 
tim's community is responding to this overt symptomatology with 
its own procedures of withdrawal, aggression, therapy, and so forth. 

What will determine the victim's and the community's expecta- 
tions of consequences and their choices of defensive strategy? Evi- 
dently the frequency, duration, and predictability of periods of 
desemantication, and their commonness in the population, will be 
data of extreme importance in the evaluation of self by the victim 
and of victim by community. If the period of desemantication is 
relatively brief (not more than a few days) , is relatively infrequent 
(not more than once a month) , is predictable (either by a calendri- 
cal device or by association with other scheduled events) , and is 
commonly observed to occur in others without dire consequences, 
then even severe degrees of desemantication with considerable asso- 
ciated inconvenience and discomfort may be tolerated by the per- 
sonality. Similarly, brief, infrequent, predictable, and common 
overt disorders may be tolerated by the community. Such situa- 
tions (to give some familiar examples) are premenstrual tension, 
drug and alcoholic intoxication, ritually induced dissociation, ex- 
haustion, and the Polar Eskimo pibloktoq. The more delayed in the 
life program, the more frequent, the more prolonged, the less pre- 
dictable, and the less common the event, the more threatening it 
will be to the personality and to the community, and the more 
desperate and (for the victim) the more ill conceived their com- 
plementary defensive strategies will become. Where the desemanti- 
cation is severe and irreversible, as in chronic brain syndromes, the 
victim may be so preoccupied with maintaining the former sense of 
competence that even trivial contretemps precipitate "cata- 
strophic" reactions (Goldstein 1940). Schizophrenia and perhaps 
the affective psychoses (such as involutional melancholia) would 
appear to have an intermediate status between chronic syndromes 
and brief episodic attacks. The desemantication is not fully contin- 
uous and the victim is consequently able to retain for a considerable 
period an intermittent normalcy of function, but the episodes are 
sufficiently frequent, prolonged, and severe to result in an accumu- 
lation of permanent defensive strategies which eventually in them- 
selves make adequate social participation almost impossible during 
the clear periods, and, sometimes, even after the desemantication 
phase itself has ended. 

But it is not merely the timing and conventionality of the dis- 
order which will affect the defensive response of the victim and 


his community. The personahty of the victim and the culture of 
the group provide models of the experiences and symptoms of the 
event which assign to them definite meanings and provide recipes 
for handling the situation. These models are, in the individual's 
case, a function of the history of his learnings, and in the com- 
munity's case, a function of other aspects of the culture, its social 
structure, and its history. They are widely variable in form and are 
not entirely predictable from a knowledge of the timing and con- 
ventionality of the disorder. While the anthropologist may or may 
not undertake the solution of problems of differential diagnosis and 
etiology (which, as we observed earlier, unavoidably involve ques- 
tions of biological as well as psychological dynamics) , he can cer- 
tainly investigate the patient's and the community's theories of 
illness and its treatment. Thus his most immediately relevant con- 
tribution can be an analysis of how, in the society in question, symp- 
tomatology and its programming are normally conceptualized. As 
we have indicated above, whatever its etiology, the course of an 
illness occurs in a social matrix and is observed both by the victim 
and his associates. Their conception of what is happening will play 
an important part in determining what will be their response to 
the symptoms (see Wallace 1959) . Thus, even if etiology and the 
primary symptoms of an illness were, except in an epidemiological 
inquiry, to be considered as physiological accidents and thus as 
largely independent of culture, the efforts of the victim and of his 
fellows to cope with the illness must be recognized as being highly 
dependent on culture, for these responses to illness are very con- 
siderably determined by what may be called the native — and, in 
particular, the patient's — theory of illness. In short, since the cause 
of illness even if physiologically initiated is progressively modified 
by feedback via the victim's and the community's conception of 
the illness, the victim's personality and the community's culture 
play a determining role. 

Some of the recent literature in social psychiatry has directed at- 
tention to theory of illness as a significant variable. Of particular 
interest are the studies of psychiatric illness in New Haven sum- 
marized in Hollingshead and Redlich's book Social Class and Men- 
tal Illness (1958). These studies demonstrate again not only class 
differentials in prevalence of certain kinds of treated mental illness 
(for example, that schizophrenia is about nine times as prevalent 
in the lowest socioeconomic group as in the highest, even after 
standardizing for population size) , but also class differentials in 



methods of treatment (that is, that lowest-class schizophrenics re- 
ceive either organic treatment or no treatment at all, while highest- 
class schizophrenics receive psychotherapy and/or organic treat- 
ment) . These differences are doubtless partly a function of 
differential access to economic resources; but, as HoUingshead and 
Redlich carefully show, they are also partly a function of differences 
in the conceptions of illness and of treatment between lower-class 
and higher-class patients. Specifically, the dissonance between the 
lower-class patients' and their middle-class physicians' theories of 
what illness is, how it originates, and how it is cured, interferes with 
free communication. These differences make mutual acceptance, 
liking, trust, and intelligent co-operation difficult, and often re- 
sult in either mutual withdrawal or the patient's refusal to enter 
into a psychotherapeutic relationship at all. 

Other sources have approached the problem of theory of illness 
from various standpoints. Cannon and others, for instance, have 
analyzed the phenomenon of "voodoo death" as a type of overre- 
sponse to a "realistically" trivial trauma by a victim who is con- 
vinced that he will die because he has been bewitched by an enemy 
or doomed for the infraction of some taboo (Cannon 1942) . Com- 
parable, if less dramatic, studies have revealed that bodily injuries 
and mental infirmities of one sort or another lead to different re- 
sponses depending on the culturally defined meaning of the situ- 
ation. For instance, in their collection of papers reporting on in- 
vestigations by the National Institute of Mental Health of the 
impact of mental illness on the family, Clausen and Yarrow describe 
in some detail the differences in the "meaning" of mental illness 
to various persons, including the patient, and the effect of these 
semantic positions in shaping the path to, through, and from the 
mental hospital (Clausen and Yarrow 1955). In their study of 
thirty-three families in which the husband was the patient, they 
found that nearly half of the husbands were never seen by a psychia- 
trist before hospitalization was arranged. The difficulty, and usually 
the reluctance, with which the patient's family came to define his 
problem as one requiring psychiatric care, and the slowness and 
uncertainty with which they proceeded to secure that care, meant 
that "discontinuities of action were frequent, and paths to the hos- 
pital were beset with obstacles and traumata for husband and wife" 
(Clausen and Yarrow 1955:32). And in our own research at the 
Eastern Pennsylvania Psychiatric Institute, we have been concerned 
with the problem of how the patient's theory of the mechanism of 


hallucination affects his and his fellows' response to that experience. 
We have worked with cross-cultural materials in the literature and 
have pointed out, for instance, the contrast between the responses 
to mescaline intoxication of normal white volunteers and of Amer- 
ican Indian religious peyotists (Wallace, 1959). 

A Model for ihe Analysis of Theories of Mental Illness 

We conceive that among the set (mazeway) of cognitive "maps" 
which each individual maintains, describing and interpreting the 
world as he perceives it, is his theory of mental illness. This map 
gives meaning to experience, by defining the possible states which 
a person can occupy in a mental health context, and by relating the 
possible states which the person can occupy to one another via vari- 
ous transfer mechanisms, so as to provide the rationale for decision. 
Such a map can therefore be conceived of as having three aspects: 
(i) the states specified; (2) the transfer mechanisms which are 
conceived to effect change from one state to another; and (3) the 
program, of illness and recovery which is described by the whole sys- 
tem. We confine our attention here to the patient's program for the 
patient himself; his programs for other persons, and the program 
of others for him, may (or may not) be different. Thus in the fol- 
lowing analyses the entity to which each state description refers 
is constant, being ego, even though ego is variable in the sense of 
having different properties at different stages of the program, and 
in the sense of being "now" at one or another of these stages in ego's 
own (not necessarily correct) opinion. (Interesting possibilities of 
programs involving multiple referent entities, because of the logical 
complexities of such schemas, are not considered here.) 

Evidently, one can "plug in" on an individual's program at a 
number of different levels of abstraction. In order to minimize 
partly the unreliability of reporting which ensues if level of ab- 
straction is left unspecified, we have found it useful to base analysis 
on five "states," which will constitute stages of every program: 
"normalcy," "upset," "psychosis," "in treatment," and "innovative 
personality." These are always to be understood as the subject's 
concepts of his own possible states and not as the observer's concepts 
of the subject's condition. The terms are unimportant; they simply 
label positions in the model. Normalcy refers to a state in which the 
person is performing to his own and other's satisfaction the roles 
appropriate to his situation in society. Upset refers to a state where 
role performance has been reduced to a level of minimal adequacy, 


with noticeable personal and/or group discomfort. Psychosis is a 
state where role performance has become so inadequate that in 
order to reduce personal and group discomfort, some degree of so- 
cial isolation (either self- or group-imposed) must be instituted. 
In treatment is a state where the person is receiving ministrations 
from specialists, designed to remove the conditions responsible for 
personal and group discomfort, and to return the patient to full 
social participation. Innovative personality is a state in which the 
person is again able to perform roles to his own and group satisfac- 
tion, but roles different to a greater or lesser degree from those 
performed in state N (as the difference approaches insignificance, 
P approaches N) . These five states may be conceived as arranged 
in a graph whose starting point is N, with "goodness" of state de- 
creasing in order of position to the right of N: 

We assume that any individual classification of states will include 
these five except where concept I is equivalent to N, in which case 
the graph reduces to: 

We also assume that between any two states one of four transfer 
relations may be conceived: no transfer possible (symbolized by 

open space); one-directional transfer ( ^); one-directional 

transfer ( ^ ) ; and reversible transfer ( * * ) . Definition of the 

states and of the transfer mechanisms can usually be best repre- 
sented not on the graph but in appended tables in order to avoid 
cluttering the graph with written notations. The reader will note 


that any two states may stand, in relation to one another, as positive 
and negative goals depending on their relative position on the value 
dimension. For instance, U may be a negative goal for a person 
who is in state N, but a positive goal for a person in state P. And 
finally, depending on the circumstances, additional states may be 
added to the model if they are part of the subject individual's or 
culture's phenomenological world. 

A given patient's theory of illness can be inferred from several 
types of behavior: 

1. Plain statements ("It's worrying that makes people lose their minds"), 

2. Comparative statements ("Joan was real sick when they brought her in, but 
now that she's been here awhile, she's quieted down a lot"). 

3. Differential motor behavior (avoiding certain patients while socializing with 
others) . 

4. Case history material (information that experiencing hallucinations first 
convinced the patient that he was seriously ill and required psychiatric help). 

These and other data, obtained from tape-recorded interviews with 
patient and his family and associates, records kept by social workers 
and therapists, direct observation on the ward, and so on, permit the 
classification of concepts and beliefs, and the working out of their 
interrelationships in the subject's mazeway. The investigator must 
keep constantly in mind that these belief structures can change and 
(this is often difficult) that it is the subject's (or the community's) 
belief system, and not the patient's "true" condition as perceived 
by the clinician, that is being studied. (And if the clinician's belief 
system is being studied, the validity of the clinician's beliefs is tech- 
nically irrelevant.) The tediousness of the task should not be under- 
estimated. A satisfactory case history, for instance, covering day- 
by-day events for months prior to hospitalization, and during the 
hospital stay itself, requires extensive checking and cross-checking 
with dozens of sources of information. The process is comparable 
to the compilation of data for a biography. Discrete items of in- 
formation, culled from various sources, are ordered first chrono- 
logically and then by topic until an internally coherent process 
appears in which the subject's decisions and attitudes are demon- 
strably related to his current situation and past experience. Thus 
one source may reveal that on a certain date the patient, a ritually 
faithful Catholic, failed to go to Mass; another source may show 
that the day before, he had an interview with his priest, who coun- 
seled him to exercise will power and to cease wallowing in self-pity; 
a third source reveals that next week the patient went to his family 


doctor and received a prescription for tranquilizers; and a fourth 
source finally shows that some time during the week preceding the 
visit to the priest, the patient experienced a frightening impulse to 
kill his wife and child. These details fit into the pattern of a process. 
With increasing fear of losing self-control, the patient, who still 
regards his "upset" state as one of moral uncertainty, turns to the 
priest for help; but the priest's advice does not help to resolve the 
uncertainty, and he redefines his state as an "illness" requiring medi- 
cal attention. 

Illustrotion: A Zulu Theory of Mental Illness 

Among the Zulu known to Canon Callaway in South Africa, 
about the middle of the last century, a complex and rather sophisti- 
cated theory was held which, in its formal structure, is not dissimi- 
lar to some varieties of current psychiatric theory. The structure of 
this theory is given in the following formula : 

N ^I) >V >F 




The definition of the states, as given in Callaway's translation of the 
Zulu text (Callaway 193 1) is as follows: 

N: "Robust"; good appetite; not choosy about food. 

D: "Delicate, not having any real disease, but delicate." 

A: "111"; choosy about food; loss of appetite; suffers vague pains; anxious 
dreams; possessed by spirits of ancestors. 

U: "111"; choosy about food; loss of appetite; suffers vague pains; anxious 

dreams; possessed by a class of spirits known as Amatongo. 
P: "A fool," "unable to understand anything," "mad/' not a "man." 

T: Continued ill health, sleeplessness, loss of weight, skin diseases, but hope- 
ful of becoming a shaman. 
S: Good physical health; the state of being a shaman or inyanga, i.e., one with 
a "soft head" who, with the help of his familiar spirits among the Ama- 
tongo, performs the respectable special role of "diviner" (finder of lost 
objects and physician to possessed persons) . 


-^ D 

D - 

-^ A 

A - 



W: "Always out of health," unable to divine, but of unusual wisdom, and 
able to work. 

The transfer operations, to the extent that they are described in 
Callaway's text, are: 

Initial possession by either Amatongo or ancestral spirits. 

Completion of possession by ancestral spirits. 

Relinquishment of possession by ancestral spirits after being 
exorcised by sacrifice of cattle under direction of shamans. 

D > U: Amatongo increase control over victim but divide into two 

groups, one group (under influence of medicines and cattle sac- 
rifice exorcism) objecting to complete possession and the other 
insisting on complete possession. 

U > P: Continued "blocking the way" of the A ;»fl/ow,^o by exorcism and 

by medicines taken by mouth. 

U > T: Patient's family, patient, and community, recognize that Ama- 
tongo are struggling to possess patient, and terminate medicines 
and exorcism. 

T > S: Patient seeks communication with Amatongo in his dreams and 

singing; community participates in his singing and ask him ques- 
tions for Amatongo to answer. 

S > W: A "great doctor" can "lay the spirit" of Amatongo to the extent 

of preventing the patient from remaining a diviner but only at 
the cost of leaving him chronically in state W. 

Notable features of the model are, first, the importance of the dif- 
ferential diagnosis (by a shaman) between possession by the rela- 
tively benevolent ancestors and by the dangerous Amatongo; and 
second, the irreversible nature of Amatongo possession, which even- 
tuates in a state of dementia unless the victim accepts his fate and 
undergoes the complete course of training as an inyanga. 

Applicafion to Clinical Case Material 

In the application of the foregoing concepts to clinical case ma- 
terial, it must be born in mind that the structure and development 
of a patient's theory of illness may be related to, but is nevertheless 
distinct from, the structure and development of his conflict struc- 
ture ("neurosis") and of his therapeutic regime. In one of the two 
cases which we have analyzed in some detail by the help of the 
model, we found the model to be helpful in understanding a tem- 
porary impasse, with an associated flurry of disturbed behavior, 
reached at a certain stage in therapy. The crucial problem in treat- 


ment, from the therapist's viewpoint, was the patient's unwiUing- 
ness to accept the presence in himself of hostile feelings toward vari- 
ous close relatives. The therapist defined the goal of treatment (I) 
as a less repressive personality and he encouraged the patient to 
assert himself and his needs more freely and to recognize that these 
needs, and the hostilities generated by their frustration, were not 
evil but merely human. The patient was stubbornly resistant, not 
merely because of the psychodynamics of the situation, but also 
because the therapist was suggesting that he "act out" in somewhat 
the same way as his own psychotic father had acted out before 
his hospitalization some years before. The therapist thus was sug- 
gesting to the patient a state I which, in the patient's theory of 
illness, was hard to distinguish from P. The patient's conscious at- 
tention was, at this time, centered on a struggle to avoid entering 
state P; hence the therapist's suggestions were terrifying, not only 
because they may have aroused unconscious resistance (in the con- 
ventional psychodynamic sense), but because they pushed him 
toward a self -identification with a psychotic father. 

The resolution of the impasse was provided by his development 
of a compromise, which the therapist was willing to accept, be- 
tween his original theory and the therapist's theory. This compro- 
mise took the following form: 

N >\] >P 


He steadfastly retained the belief that the object of his efforts was 
a return to his normal, presymptomatic, good-husband-and-f ather 
self (N) . But he accepted T as a necessary way station on the path 
to N and as a means of avoiding the alternative state P. His ac- 
ceptance of the existence and value of T were followed almost im- 
mediately by release to the outpatient department. 

Application to the Classification of Cultures 

Because of the ubiquity of the major types of mental disease, 
and because of the uncertainty of etiological understanding, it is 
hazardous to classify cultures as more or less pathogenic in respect 
to any particular mental illness or to mental illness in general. In all 
likelihood, as knowledge of the causes of mental illness is extended, 
it will become easier to discern the relation between culture and 


etiology. Thus in the future it may be possible to regard the fre- 
quency, distribution, and forms of mental illness in a society as an 
index of its culture. But at the present time, despite the currency 
of certain hypotheses based on psychodynamic assumptions about 
the relation between culture and mental illness, it is not feasible to 
establish a classification based on demonstrated etiological processes. 

It is however reasonable to suggest that cultures may, even on the 
basis of present knowledge, be classified with respect to such cul- 
turally institutionalized responses to various types of mental illness 
as the society's taxonomy and definitions of mental illness, its theory 
or theories of illness, and its techniques of therapy and their ra- 
tionale. Such a classification must, in effect, form a matrix of in- 
tersection of a constant typology of mental illness (that is, a typol- 
ogy defined by the investigator and used as a constant referent for 
controlling cross-cultural comparisons) and of alternatively pos- 
sible responses available cross-culturally. The types so defined may 
then be investigated in order to discern whether or not a correlation 
exists between response type and other aspects of culture. If such 
correlations can be shown to exist, then at least response to mental 
illness may be considered an index of culture. 

Evidently a number of possible schemes, of varying degrees of 
complexity and abstraction, can be created, based on different con- 
stant typologies and different panels of alternative responses. One 
typological system based on theoretical considerations introduced 
in the preceding sections will be outlined here. For the constant 
typology, not Western diagnostic categories, but the two dichoto- 
mous dimensions of severity and chronicity will be used (mild 
versus severe, and intermittent versus continuous) . For the response 
typology, two dichotomous dimensions will be used: episodic versus 
symptomatic interpretations of illness, and treatment versus ex- 
trusion as a method of handling illness. These concepts may be de- 
fined further as follows: Mildness and severity refer to the degree 
of abnormality of the overt behavior itself and not to its duration 
or frequency of occurence; intermittency and continuousness refer 
to halves of a continuum, intermittency being the half in which 
the disorder can best be characterized as discrete attacks separated 
by intervals of normalcy, and continuousness as the half in which 
the disorder can be characterized as a period of uninterrupted dis- 
function. Episodic interpretations of illness confine attention only 
to the overt disorder itself and regard it as an isolated episode in an 
essentially normal life program, whereas symptomatic interpreta- 


tions construe the overt disorder as a sign of a more serious under- 
lying inadequacy which threatens to recur, possibly in a more un- 
desirable form, on later occasions. Treatment as a method of 
handling illness implies a policy of attempting to cure, to improve, 
or to tolerate (even by ignoring the behavior) and make the best use 
of the victim, in contrast to the method of extrusion, which by 
such devices as confinement, banishment, or even execution at- 
tempts to rid society entirely of an incompatible participant. The 
suggested dichotomies are, of course, divisions of continua, and the 
distinctions are easier to make in extreme than in intermediate cases. 
Thus a series of epileptic attacks is easy to classify in the constant 
typology as intermittent and severe, and a case of obsessive fear of 
heights as mild and continuous ; but a given schizophrenic psychosis 
may be neither clearly continuous nor notably severe, yet seem by 
contrast with epilepsy and the fear of heights to require the con- 
tinuous and severe classification. 

The whole schema may be represented in the following diagram: 





Episodic or Symptomatic 
Treatment or Extrusion 

Episodic or Symptomatic 
Treatment or Extrusion 

Episodic or Symptomatic 
Treatment or Extrusion 

Episodic or Symptomatic 
Treatment or Extrusion 

Thus any group, with respect to any given syndrome, may be classi- 
fied as episodic-treatment, episodic-extrusion, symptomatic-treat- 
ment, or symptomatic-extrusion, within that cell which character- 
izes the syndrome on the constant typology. If we consider 
pibloktoq, for instance, we would classify this as intermittent-se- 
vere in the constant typology, and the Polar Eskimo handling of 
it as episodic-treatment in the response typology. The same syn- 
drome in the context of, let us say, an operational wing of the U.S. 
Strategic Air Command would also be classified as intermittent- 
severe, but the handling of the condition would be classified as 
symptomatic -extrusion. And, again, this same intermittent-severe 
syndrome in the context of a liberal arts college campus would be 
handled either as episodic-treatment or symptomatic-treatment. 

The number of possible cultural patterns established by this 
paradigm is quite large. Although, with regard to any single syn- 



drome, only four types of response are considered, there are four 
types of syndrome, with regard to each of which these four possi- 
bihties exist. Therefore the number of possible cultural patterns 
is 4^ or 256. Furthermore, of course, any description of the way in 
which a society handles mental disorders will make many distinc- 
tions, even of a classificatory kind, that cannot be included in a 
pattern classification scheme. Thus, for instance, with respect to 
the "treatment" class, it will be noted in any description whether 
the condition in question is ignored, is recognized but tolerated, or 
is directly approached by a means of therapy. If therapy is em- 
ployed, it can be medical (physiological) or psychological; and if 
psychological, it can be secular or religious, cathartic or repressive, 
and so on. Rather than attempt to embrace all of the 256 patterns, 
let alone the further elaborations and refinements desirable for any 
sort of descriptive account, therefore, it would appear to be useful 
to note that among the large number of possible patterns, several 
stand out as stock patterns which may be used for the purpose of 
seeking to establish whether or not, in principle, correlations may 
exist between a group's manner of handling behavior disorder and 
other aspects of its culture. 

Four such ideal pattern types are offered below: 













































It is suggested' — with the hope not so much that the suggestions will 
convince as provoke thought and consideration in empirical studies 
— that these four patterns of institutionalized response to mental 
illness are associated with definite types of social structures. Pat- 
tern I, for instance, would seem to be characteristic of aggressive 
and power-seeking, self-selected, elite groups generally, whether 
they be kinship, military, political, economic, or religious. These 
elite groups extrude (screen out) all persons with visible behavioral 
anomalies (symptomatic of possible other disabilities as yet unre- 
vealed) in order to maintain a maximally reliable and effective or- 
ganization. Pattern II would seem to be characteristic of techno- 


logically primitive, small communities that recognize disorder as a 
symptom of a hidden, threatening weakness only when it is con- 
tinuous, and that will resort to extrusion only when it is both con- 
tinuous and severe. Pattern III would seem to be characteristic of 
prenineteenth century Western civilization generally: all disorders 
are symptomatic, and all serious disorders require extrusion. Pattern 
IV, on the other hand, would seem to characterize the psychody- 
namic tradition in twentieth century Western psychiatry, and an 
increasing number of other educated subgroups in Western popu- 
lations, who regard all disorders as symptomatic, but also consider 
that all disorders should be treated rather than disposed of by ex- 

Space does not permit further elaboration of these concepts; but 
enough has been said, perhaps, to indicate not only the problems in 
attempting to create a taxonomy of responses to mental illness with 
cultural index value, but also the possible value of such a taxonomy 
in establishing relations between responses to mental illness and 
other aspects of culture. To the extent that these patterns of re- 
sponse have a bearing on the course of various syndromes, whatever 
their etiology may be, a taxonomy of this kind may additionally 
have some utility as an evaluative index of social efficiency in han- 
dling the problems of mental illness. We may speculate, for in- 
stance, that a group whose response to a behavioral disorder is to 
regard it as symptomatic of an underlying and threatening chronic 
incompetency, rather than an episode in a normal life program, will 
induce in the victim a sense of his own inadequacy that is in itself 
directly pathogenic. We may further speculate that his anxious 
efforts to defend himself will markedly affect the form and course 
of the disorder itself. If these defensive efforts are not directed 
toward the securing of a validly effective therapy, then the patho- 
genic pressure of the culturally institutionalized definitions of 
and responses to mental illness will be uncompensated. In such an 
unhappy case, even if the etiology of the disorder were actually 
completely organic, the culture would be playing a contributory 
role in the mental disease process. 




How can the cultural anthropologist relate his conceptions of 
the structuring of social behavior to biological theories of mental 


illness? The model of mental illness advocated in this paper as an 
answer to this question is essentially homeostatic. A behavior sys- 
tem is considered to be disturbed when an independent variable, 
organic in nature, passes certain boundary values; and the responses 
of the various components of this system can be construed as moti- 
vated efforts to restore equilibrium. These responses are prescribed 
by the system itself in its theory of illness. But mere lip service to 
the ideal of an "interdisciplinary" approach, and pleas for the recog- 
nition of the importance of biological or cultural factors, will not 
solve the scientific problem. Only an approach which considers the 
specific nature of the interaction between biological and cultural 
(psychosocial) variables can have high predictive value. 

The specific nature of this biocultural interaction can best be 
investigated by conceiving of the total course of the psychotic 
episode as a single event and then analyzing it into stages. Each stage 
is defined by a change in one of the major relevant dimensions of 
the event. A number of plausible programs can be constructed by a 
priori reasoning from different assumptions about the identity of 
the initial stage. One such program derives from the assumption 
(not yet justified by empirical findings) that the initial event in 
the psychotic episode is the occurrence of an organic disfunction 
in a hitherto intact (even if peculiarly vulnerable) individual. 

If one makes this assumption, every episode of serious mental ill- 
ness can be divided into four stages (exclusive of therapeutic and 
rehabilitation stages) . 

In the first stage, the organism is functioning normally. 

In the second stage, an intermittent or continuous, of greater or 
lesser severity, organic interference with normal brain function oc- 
curs. Presumably the oft-remarked transcultural invariance of the 
major clinical entities and the absence of unique ethnic psychoses 
result because the number of types of organic interference is lim- 
ited. Many sources of such interferences are known, however: 
cerebral hypoglycemia or hypoxia, electrolyte disturbances, gross 
tissue change, hormonal autointoxication, toxic metabolites, drugs, 
viral invasion, anomalies of enzyme action, and so on. These im- 
mediate sources in turn can theoretically depend upon many "final" 
causes, including prolonged states of psychodynamically and so- 
cially determined stress (such as those revealed by psychoanalytic 
investigations) which may produce temporary, and conceivably 
sometimes even irreversible, changes in body chemistry. Genetic 
factors may also be responsible for differential vulnerabilities within 


a population to the various noxious factors. Thus even from an 
organismic position one can comfortably look to social and psycho- 
logical processes as "final" causes, particularly if the differential 
incidence of disorders rather than the understanding of individual 
cases is of primary concern. Coincident with the neural dysfunc- 
tion occurs psychological dysfunction. The quality of this dysfunc- 
tion is best conceived as a relative difficulty in organizing cognitive 
content: difficulty in finding the "meaning" of perceptual data, 
difficulty in maintaining the structure of motives, difficulty in re- 
lating affect to "rational" considerations. These difficulties may be 
metaphorically described as desemantication: the shrinking of the 
semantic matrix. This kind of dysfunction can vary in severity 
from an almost imperceptible decrement to a decrement so cata- 
strophic as to approximate decerebration, with attendant loss of 
perceptual contact with the environment, motor discharge, and 
release of autonomic functions. At an intermediate level between 
mild confusion and unconsciousness would seem to fall the experi- 
ence of meaninglessness, described by some schizophrenics as a sense 
of unreality, depersonalization, and loss of identity. Desemantica- 
tion may be briefly episodic, as in hysteriform attacks, or chronic, 
as (apparently) in schizophrenia. Also coincident with neural and 
psychological dysfunction is primary behavioral failure attendent 
upon the desemantication. This is failure as judged by either the 
victim and members of his group, or both, and may occur in a 
variety of sectors of life, both interpersonal and technological. 
While incompetence in interpersonal relations may be the most 
conspicuous consequence of desemantication in the eyes of the 
group, technical failures in performing essential routine tasks, such 
as walking, paddling a kayak, ironing clothes, and preparing food, 
may come first to the victim's own awareness. Such failures may 
vary in duration and in the social or individual importance of the 
area of behavior involved. 

If negative self -evaluation by the victim follows the events of the 
second stage, then the third stage will occur, characterized by 
anxiety, depression, and other negative affects directed toward the 
self. All persons constantly monitor and evaluate their competence 
in attaining their goals, both by self-perception and by perception 
of others' response to their behavior. A person experiencing dese- 
mantication finds the performance of his tasks more difficult and 
in some instances impossible. If the desemantication is continuous 
and is relatively severe, he will be unable to deny the reality of his 


loss of competency. His evaluation of these failures, which is a 
complex function of his current experience, the responses of others, 
and past learning, will be less effective than normal precisely because 
of the desemantication itself. But it will be based, in every instance, 
in part on concepts available to him from his past learning of the 
culturally standardized interpretations of the specific experiences 
and incompetencies which he now recognizes in himself. Thus he 
may interpret the perplexing voices which he hears as religious reve- 
lations, as the delirium accompanying fever, as the result of over- 
work, as the consequence of emotional conflict, and so forth, de- 
pending on the content of the experience, the reactions of others, 
and the explanations offered by his own cultural background. To 
the extent that the self-evaluation is negative, he loses confidence 
in his ability to control his own behavior, to master his environment, 
and to relate his behavior systematically with others. 

The fourth stage is cognitive damage incurred in the course of 
the victim's defensive response to the negative self-evaluation. The 
response to his own anxiety and depression is, because of the exist- 
ence of physiological dysfunction, itself apt to be disorganized. But 
it is designed to improve the negative self-image and to protect the 
person from catastrophe, and may in some degree relieve the pa- 
tient's anxiety and depression, albeit at the cost of cognitive damage 
in the form of paranoid delusions, self-limiting withdrawal from 
society, and so on. Part of the response may be "neurotic," in the 
sense of utilizing such mechanisms of defense as denial, repression, 
projection, paranoid oversimplification, and so on. Part of it may 
be impulsive fighting with, or withdrawing from, a now dangerous 
and exhausting world. Part of it may take the form of seeking help. 
The style in which the person goes about attempting to defend him- 
self, maintain self-respect, and secure help will of course reflect his 
cultural learning. 

Through the second, third, and fourth stages, the victim's com- 
munity is also evaluating and responding to him as a "changed per- 
son." Even in a homogeneous community, the social evaluation and 
response may be considerably different from the victim's, both be- 
cause the victim's desemantication constrains his behavior, and be- 
cause his motives may be divergent from those of the group. 
Whether or not his motives diverge from the group will depend 
considerably on the nature of these beliefs. Thus, for instance, if 
mental illness as evidenced by hallucination is culturally defined as 
a degrading condition to which society responds by social extrusion, 


the victim will be strongly motivated to conceal his condition, 
to deny it, to withdraw from prying eyes, and to accuse others of 
conspiracy against him if the charge is made. If, on the other hand, 
hallucination is a sign of contact — uncomfortable perhaps — with 
the supernatural world, and is responded to with rituals of intensi- 
fied social acceptance, the hallucinator's motives will in all likeli- 
hood not be directed toward denial, concealment, and defense, but 
toward maximum publicity. 

This model of the process of becoming mentally ill, as an imme- 
diate consequence of neurophysiological dysfunction, in a social 
environment, may be succinctly represented in a paradigm. Such a 
paradigm, of course, represents only a canonical form or modal 
type. The symbols are read as follows: "O" represents level of 
neurophysiological function of brain; "S" represents level of se- 
mantic psychological function; "B" represents level of overt be- 
havioral success in achieving goals in social context; "A" repre- 
sents level of anxiety, depression, and other negative affect directed 
toward self; and "D" represents the degree of cognitive damage 
incurred in the course of the defensive responses of the individual 
to his own negative self-evaluation. The operator j represents 
pathological change, and A represents "and." 

Stage o: Eufunction (0,S,B,) A (A) A (D) 
If physiological injury occurs, then 

Stage i: Primary Dysfunction (jO, jS, jB) A (A) A (D) 
If negative self -evaluation occurs, then 

Stage 2: Anxiety and Depression ( | O, J, S, | B) A ( i A) A (D) 
If anxiety and depression are severe and prolonged, then 

Stage 3: Cognitive Damage ( | O, | S, j B) A (A) A ( j D) 


The importance of the organic factors in psychopathology has 
been largely ignored by anthropological theory, which has empha- 
sized psychological factors almost exclusively. If the viewpoint is 
taken that organic events play a significant role in the etiology of 
many mental disorders, it is possible to see the role of cultural dif- 
ferences as particularly relevant to etiology via their influence in 
determining the frequency with which the pathogenic organic 
events occur. From this point of view also, the culturally institu- 
tionalized theories of illness and of therapy appear to be extremely 
important in deciding the nature of the victim's and his group's 


responses to the disorder. A model of mental illness as a type of event 
is offered which integrates the organic and psychosocial approaches. 
It may be hoped that anthropologists who have occasion to make 
observations in the field on persons with mental illness will in the 
future be able to obtain and record more extensive information on 
the physical status and history of the victims. Data on nutrition, 
infectious diseases, head injuries, and autonomic symptomatology, 
both with regard to the individual cases and also with respect to 
the community as a whole, would be helpful in describing indi- 
vidual cases, in understanding group differences, and in putting the 
brakes on overly facile attributions of psychopathology to "social 
structure," "culture," and "basic personality." 


American Museum of Natural History, Photographic Division. 

1 9 14 Photographs of the Crockerland Expedition. 
Baashuus-Jensen, J. 

1935 Arctic nervous diseases. Veterinary Journal (London) 91:339-350, 
Barrett, Albert M. 
1919-20 Psychosis associated with tetany. American Journal of Insanity 76:373- 

Bellak, Leopold (ed.) 

1958 Schizophrenia: a review of the syndrome. New York, Logos Press. 
Bertelsen, a. 

1940 Gr0nlandsk medicinsk statistik og nosografi. Meddelelser om Gr0nland, 
Bd 117, Nr. 3. Copenhagen. 
Brill, A. A. 

19 1 3 Pibloktoq or hysteria among Peary's Eskimos. Journal of Nervous and 
Mental Disease 40:514-520. 
Callaway, Canon H. 

193 1 The religion of the Amazulu of South Africa, as told by themselves. In 
A. L. Kroeber and T. T. Waterman, Source book in anthropology. New 
York, Harcourt, Brace. 
Cannon, Walter B. 

1942 Voodoo death. American Anthropologist 44:169-181. 
Clausen, J. A., and M. R. Yarrow 

1955 The impact of mental illness on the family. Journal of Social Issues 
11: (4) (whole issue) . 
Cook, Frederick A. 

1894 Medical observations among the Esquimaux. Transactions of the New 
York Obstetrical Society, 1 893-1 894, pp. 171-174. 

* A short bibliography of other works dealing with the same subject but not referred to 
here is appended at the end of this chapter. A selected bibliography bearing on the mutual relation- 
ship between anthropology, psychiatry and psychoanalysis is given in the Appendix at end of the 



1904 Tetanic, psuedotetanie und ihre mischformen bei hysteric. Deutsche 
zeitschrift fiir ncrvenhcilkundc 27: article 12, 239-268. 

Glass, Albert J. 

1953 Psychotherapy in the combat zone. In Symposium on Stress, Washing- 
ton, Walter Reed Army Medical Center. 

Goldstein, Kurt 

1940 Human nature. Cambridge, Harvard University Press. 

Gussow, Z. 

i960 Pibloktoq (hysteria) among the Polar Eskimo: an ethnopsychiatric 
study. In W. Muensterberger, ed.. Psychoanalysis and the Social Sciences. 
New^ York, Ruternational Universities Press. 

Hollingshead, a. B., and F. C. Redlich 

1958 Social class and mental illness. New York, Wiley. 


1 94 1 Studies on the nutrition and physio-pathology of Eskimos. Oslo, 
Skrifter utgitt au Det Norske Videnskaps-Akademi i Oslo, I. Mat.- 
Naturv. Klasse 1940 No. 9. 

Hunt, J. McV. (ed.) 

1944 Personality and the behavior disorders. Nev/ York, Ronald Press. 

Kallman, Franz 

1938 The genetics of schizophrenia. New York, J. J. Augustin. 

Kane, E. K. 

1856 Arctic explorations: the second Grinnell expedition. Philadelphia, Childs 
and Peterson. 

MacMillan, Donald B. 

19 1 8 Food supply of the Smith Sound Eskimos. American Museum Journal 

18: 161— 176. 
1934 How Peary reached the pole. Boston, Houghton. 

Mandelbaum, David G. 

1949 Selected writings of Edward Sapir. Berkeley, University of Cahfornia 

Maxwell, J. P. 

1930 Further studies in osteomalacia. Proceedings of the Royal Society of 
Medicine 23:63 9-640. 

MiLBANK Memorial Fund 

1952 The biology of mental health and mental disease. New York, Hoeber. 

Peary, Robert E. , 

1907 Nearest the pole. New York, Doubleday, Page. 
Peterson, Donald B., et al. 

1950 Role of hypnosis in differentiation of epileptic from convulsive-like 
seizures. American Journal af Psychiatry 107:428-443. 

Rasmussen, Knud 

1915 Foran Dagens 0je: Liv I Gr0nland. Copenhagen. 



1957 Human acclimatization to cold. Arctic Aeromedical Laboratory, Tech- 
nical Report 57-21. 

Selye, Hans 

1956 The stress of life. New York, McGraw-Hill. 

Shelling, D. H. 

1935 The parathyroids in health and disease. St. Louis, Mosby. 

Tooth, Geoffrey 

1950 Studies in mental illness in the Gold Coast. London, H. M. Stationery 

"Wallace, Anthony F. C. 

1959 Cultural determinants of response to hallucinatory experience. A.M. A. 

Archives of General Psychiatry 1:58-69. 
1959 The institutionalization of cathartic and control strategies in Iroquois 
religious psychotherapy. In Marvin Opler, ed., Culture and mental 
health. New York, MacMillan. 
Whitney, H. 

191 1 Hunting with the Eskimos. New York, Century. 




Arieti, Silvano, ed. 

1959 American handbook of psychiatry. 2 vols. New York, Basic Books. 
Bellak, Leopold, ed. 

1958 Schizophrenia: a review of the syndrome. New York, Logos Press. 
Best, Charles H. and Norman B. Taylor 

1955 The physiological basis of medical practice. 6th ed. Baltimore, Wi-Uams 
and Wilkins. 

Davidson, S., A. P. Meiklejohn, and R. Passmore 

1959 Human nutrition and dietetics. Baltimore, Williams and Wilkins. 
Dewan, John G. and William B. Spaulding 

1958 The organic psychoses. Toronto, University of Toronto Press. 
Duncan, Garfield G., ed. 

1952 Diseases of metabolism. 3d ed. Philadelphia, W. B. Saunders. 
Goldstein, Kurt 

1940 Human nature. Cambridge, Harvard University Press. 
HosKiNS, R. G. 

1946 The biology of schizophrenia. New York, W. W. Norton and Co. 
Kallman, Franz 

1938 The genetics of schizophrenia. New York, J. J. Augustin. 
Kline, Nathan S. 

1956 Psychopharmacology. Washington, D.C., American Association for the 
Advancement of Science. 


Merritt, H. Houston and Clarence C. Hare, eds. 

1953 Metabolic and toxic diseases of the nervous system. Baltimore, Williams 
and Wilkins. 

MiLBANK Memorial Fund 

1952 The biology of mental health and mental disease. New York, Hoeber. 
Pfeiffer, John 

1955 The human brain. New York, Harper & Bros. 

Research Publications of the Association for Research in Nervous and 
Mental Disease. 

Sargant, "William 

1954 An introduction to physical methods of treatment in psychiatry. Balti- 
more, Williams and Wilkins. 

Selye, Hans 

1950 The physiology and pathology of exposure to stress. Montreal, Acta. 

1956 The stress of life. New York, McGraw-Hill. 
Williams, Roger J. 

1956 Biochemical individuality. New York, Wiley. 

chapter i o 



Harvard University 

Historically, the investigation of dreams has occupied an inter- 
esting position in anthropological theory. From an early position of 
prominence in the nineteenth century, the study of dreams became 
more and more peripheral to the major interests of anthropolo- 
gists. Perhaps this was due to the Freudian revolution, which radi- 
cally altered the general conception of dreams, and also to the shift 
in interest away from cultural evolution. Before Freud, dreams 
had been considered a possible major influence on the origin and 
development of religion. After Freud, dreams came to be considered 
disguised representations of motives, and, therefore, more relevant 
to the analysis of the individual psyche than to an analysis of social 
and cultural events. 

However, by 1930, anthropologists had begun to raise questions 
concerning the psychoanalytic theory of dreams, and especially the 
assumption that dream symbols had the same meaning in societies 
with cultural traditions very different from those of Western 
Europe. Since 1930, numerous dreams from non-Western peoples 
have been interpreted by field workers interested in the relation 
between culture and personality, and the psychological uses and 
functions of dreams in a number of non-Western societies have also 
been examined. Thus, dreams have again entered into the discus- 
sion of man's cultural life, although in a psychological rather than 
evolutionary context. 

This review will attempt to bring together the anthropological 
findings concerning dreams, focussing on the interaction of cul- 
tural and psychological factors. First, as a historical introduction, 
the influence of dreams on the origin and development of culture 



will be discussed. Next the question of universal symbolism in 
dreams will be considered, followed by a review of the results of 
psychological interpretations of dreams from "primitive" or non- 
Western societies. The last two sections will consider the ways in 
which culture influences and utilizes dreams. 

The Influence of Dreams on fhe Origin and Development of Culture 

The cultural study of dreams begins in modern anthropology 
with Tylor's work on animism. Tylor considered animism, or the 
"doctrine of the soul," to be the basic substratum of religion, from 
which arose more complex forms of religious belief. In order to ac- 
count for the origin of animism, Tylor turned to the ethnographic 
materials available to him concerning dreams, and especially the 
widespread belief that during dreams the soul may travel about, 
meet other souls, and receive injuries which may affect the health of 
the dreamer. From this association of dreams and beliefs about the 
soul, Tylor inferred that the idea of the soul arose from man's at- 
tempt to account for the phantom visitors in sleep and dreams, and 
to explain the differences between sleeping and waking, and life and 
death. This thesis has been found plausible by a wide range of schol- 
ars. Lowie stated: 

His (Tylor's) theory is avowedly a psychological interpretation pure and 
simple, but inasmuch as it not only explains the empirical observations, but oper- 
ates exclusively with facts like death, dreams, and visions, all of which demon- 
strably exercise a strong influence on the minds of primitive man, it must be 
conceded to have a high degree of probability. I for one, certainly have never 
encountered any rival hypothesis that could be considered a serious competitor. 
( 1924:108) 

Tylor was not the first to present this idea. Thomas Hobbes, 
in 1 65 1, stated: 

From this ignorance of how to distinguish Dreams and other strong fancies 
from Vision and Sense did arise the greater part of the religion of the Gentiles 
in times past that worshipped Satyres, Faunes, Nymphs, and the like; and nowa- 
days the opinion that rude people have of Fayries, Ghosts, and Goblins, and the 
power of Witches. (Leviathan, ch. xii, quoted by Jones 193 i.) 

A further elaboration of this thesis from the psychoanalytic point 
of view has been presented by Ernest Jones, who postulates that 
the conceptions of werewolves, vampires, incubi, witches, and the 
devil, current in the Middle Ages, were also derived from dream 
experiences. Specifically, Jones tries to show that these supernatural 
figures were derived from nightmares, with which they share nu- 


merous common features, such as an identical latent content repre- 
senting incestuous wishes, extreme dread, transformation of per- 
sons into animals, the occurrence of fantastic animal forms, the 
alternation of the imagined object between extreme attractiveness 
and intense repulsiveness, the idea of flying through the air, and the 
representation of sexual acts as torturing assaults. Also these crea- 
tures appeared frequently in nightmares, and were, at the time, 
considered to be a direct cause of nightmares (1931:239). 

Lincoln, in the Dream in Primitive Culture, presents a number 
of ethnographic examples in which cultural items, such as curing 
rituals, art work, songs and dances, religious cults, and so forth, 
were supposedly invented in dreams, Lincoln points out that it is 
usually in culturally defined and expected dreams that these cul- 
ture items originate, rather than idiosyncratic dreams, and that 
these culture items are often presented in the dream by an ancestor- 
like spirit. Lincoln concludes : 

... a large part of primitive culture is a result of the dream, or more accurately 
a result of the psychological and cultural processes behind the dream. These 
processes are given form in the dream and influence the culture directly from the 
latter. (1935:93) 

The general thesis of Tylor, Jones, Lincoln, and others that 
dreams have been either a primary or secondary source of innova- 
tion is a difficult argument to prove or disprove. Generally, the 
argument is based on similarities between typical dream experiences 
and cultural items and the frequent appearance in dreams of these 
items. Even where there are cultural demands that an individual in- 
vent a song or myth or ritual in dreams, this process is often only a 
minor reworking of already existing materials, for which the dream 
is as much an expected means of validation as an actual source of 
invention. Devereux, in a careful study of Mohave dreams and 
rituals, finds: 

Although Mohave shamans and singers are supposed to acquire their knowledge 
in dreams they actually learn it in waking Hfe and then have dreams which con- 
dense or allude to this body of knowledge. (1957:1044) 

While the influence of dreams on the origin and development of 
culture remains obscure, these studies have clarified one aspect of 
the relation of dreams to culture. They document the persistent 
association found between dreams and beliefs about supernatural, 
including other souls. Dreams have been shown to be one of the 
chief means of communication with supernaturals, and super- 


naturals have been found to have certain similarities to figures 
which typically appear in dreams. It appears that dreams, as a com- 
mon projective experience, merge with and take material from 
other culturally defined projective systems, and also give the indi- 
vidual direct access to these projective systems, although the degree 
of merger and access varies from culture to culture. 

Universal Symbolism in Dreams 

Of the complex theory presented by Freud in the Interpretation 
of Dreams, anthropologists have discussed most frequently the 
problem of the universality of dream symbols. The relevance of this 
problem has been succinctly put forward by C. G. Seligman. 

If it can be shown that identical symbols (i.e., identical symbols with the same 
meaning attached to them) prevail, then we shall have to admit that the uncon- 
scious of the most diverse races is qualitatively so alike that it actually constitutes 
a common store on which fantasy may draw, and it becomes imperative to give full 
weight to this in any discussion of the origin of myths and beliefs. (1927:200) 

In the Interpretation of Dreams, symbolization is treated as just 
one of the processes by which the latent content, consisting of a 
series of thoughts expressing a wish, may be transformed into the 
jumble of vivid images which comprise the manifest content of the 
dream. The latent dream thoughts may also be changed through 
condensation and displacement, that is, by the use of hints or 
allusion, or by the substitution of a part for the whole, or by repre- 
senting words in pun-like images. Unlike the processes of conden- 
sation and displacement, however, symbolization is considered to 
be more limited in scope, and to refer, especially in dreams, to only 
a limited number of things: the nuclear family, the human body, 
and the biological activities of the body, such as birth, sucking, def- 
ecation, copulation, and death (Freud 1920:156-177). 

"While the hypothesized processes of condensation and displace- 
ment have not been subject to much question, the possibility that 
man uses a special vocabulary without awareness and tuition has 
been the subject of frequent debate. And the possibility that this 
vocabulary is everywhere the same, impervious to culture, has also 
raised questions, especially among anthropologists. It should be men- 
tioned that the issue here is not whether dreams can be interpreted 
solely through symbols, but whether a certain cognitive process, 
symbolization, is universal. Any interpretation of dreams based on 
symbols alone would leave out all the other processes of dream for- 
mation, thereby omitting the analysis of a large amount of material. 


and resulting in an oversimplified if not incorrect analysis (Freud 
1900:353, Eggan 1952, Roheim 1947:88). 

Unfortunately, actual investigations of dream symbolism in non- 
Western cultures have been relatively rare, perhaps because of the 
great methodological difficulties. Those studies of non-Western 
dreams which have used symbols in dream interpretation have some- 
times brought out impressive convergences with other personality 
data (Devereux 1951, Lee 1958). However, these convergences 
cannot be considered good evidence for the universality of dream 
symbols, since the conclusions derived from the interpretation of 
dream symbols have not been compared systematically with, nor 
constructed independently from, other personality data. 

There is, however, some assessable evidence on the issue of sym- 
bolism in dreams. Within Western culture, supportive evidence has 
been reported from the studies of hypnotic dreams, in which a 
subject is instructed to dream of certain activities in a hidden or dis- 
guised way, and to remember the dream but to forget the instruc- 
tions upon waking. Roffenstein reports the following dream by an 
uneducated woman told to dream of sexual intercourse with her 

I dreamt about my father, as if he had presented me with a great bag, a traveUng 
bag, and with it he gave me a large key. It was a very large key. It looked like 
a key to a house. I had a sad feeling, and I wondered about its being so big; it 
couldn't possibly fit. Then I opened the bag. A snake jumped out right against 
my mouth. I shrieked aloud and then I woke. (1951:255) 

The symbols of the key and the snake for the penis, and the traveling 
bag and mouth for the female genitalia, stand out clearly. While 
this kind of evidence supports the symbolism hypothesis, one posi- 
tive instance is not a proof. 

However, even if Western dreamers do use (sometimes, at least) 
a stereotyped set of symbols, it would be rash to then assume that 
these same symbols are used in the same way in all cultures. One ex- 
cellent but laborious method of investigating the universality of 
dream symbolism is to compare interpretations based on the sym- 
bolism found in the dreams of non-Western informants with iude- 
pendently collected history materials. An example of this method 
has been reported by Honigman. 

It has been suggested that in dreams protrusions symbolize the male sex organ 
and the male's normally assertive role in copulation. Aware of this cUnically de- 
rived interpretation, we implicitly predicted sexual inadequacy or impotence for 
a young Kaska man who reported a dream in which he was attacked by a grizzly 


bear. "My gun stick. I get nervous. I try to take shot at him. My gun got no 
power. Goes ssssss — goes out quick." When interviewed, the informant rejected 
the interpretation equating gun and penis. Our prediction was nevertheless con- 
firmed when on two subsequent occasions, according to reUable testimony, the 
informant experienced acute impotence." (1954:158-159) 

Another kind of evidence for the universahty of dream symbols 
is found in the meaning that particular cultures assign to certain 
dreams. It has been found that a number of cultures, widely dis- 
persed, attribute to certain common dreams similar meanings which 
correspond closely to the psychoanalytic interpretations of these 
dreams. C. G. Seligman has collected a number of examples of such 
similarities (1924, 1932), and a review of this material may be 
found in Lincoln (1935:107-131). Two of the most remarkable 
of these similarities are the interpretations that feces in a dream 
stand for wealth, which is reported for the Ashanti, Tikopia, West- 
ern Europeans, Thai, Tangerians, Naga, Chinese, and Sinhalese, and 
the interpretation that loss of a tooth in a dream indicates death, 
illness, or disaster, reported for the Lolo, Araucanians, Chuckchee, 
Western Europeans, Chiricahua, Cuna, Ashanti, Naga, Malayans, 
Achelenese, Japanese, Chinese, and Diegueno (Seligman 1924) . Al- 
though these interpretations are not made by every culture, it seems 
unlikely that widely dispersed cultures should have hit upon such 
similar symbolism by chance. 

From these bits of evidence, it would seem that some degree of 
universal symbolism in dreams is probable. Seligman's conclusions, 
set down in 1924, seem to be still adequate. He stated: 

The essential dream mechanisms of non-Europeans including savage and bar- 
baric peoples, appear to be the same as in ourselves. Thus dreams with symbolism, 
sometimes elaborate and recondite, often simple and obvious, occur. These dreams 
may be wish-fulfillments or be provoked by conflict. 

Dreams with the same manifest content to which identical (latent) meanings 
are attached (type dreams) occur, not only in cognate groups, but among peoples 
of diverse race and in every stage of culture. (1924:46) 

A complete validation of the hypothesis of universal symbolism 
is, of course, impossible. The important issue is the degree of proba- 
bility which is to be assigned to this hypothesis. Perhaps it would be 
more fruitful to investigate the degree to which universal symbols, 
in dreams and in other fantasy materials, can be laid over with 
secondary cultural and individual meanings, and the degree to 
which culture is selective in choosing from the stock of possible 
symbols, than to try to document such an unwieldy issue as the uni- 


versality of symbols. Lee, for example, in his study of Zulu dreams, 
demonstrates the way in which different dream symbols are used 
by women at various points along the life cycle. 

Lee finds that unmarried women are more likely to have symbolic 
birth dreams of still water, compared to married women with few 
children, who are more likely to have undisguised dreams of babies. 
Married women with -many children, however, are likely to have 
frightening symbolic birth dreams of flooded rivers (1958). This 
selectivity in dream symbolism supports the hypothesis that symbols 
are less likely to be used if the wish symbolized is acceptable, and also 
demonstrates something of the complex relations between cultural 
norms, social roles, and motivation. 


This section will review some of the findings and basic issues 
involved in the psychological interpretation of dreams from non- 
Western societies. The theory and techniques of dream interpreta- 
tion used by field workers in non- Western societies have been based 
on Freud's monumental Interpretation of Dreams, although a num- 
ber of warnings about complete acceptance of the psychoanalytic 
methods and theory have been presented by anthropologists who 
have worked with non- Western dreams (Eggan 1952, Honigmann 


From a scanning of the published interpretation of non- Western 

dreams by Lincoln ( 1935) ,Roheim (1946, 1949, 1950) ,Devereaux 
(1951), and Kluckhohn and Morgan (1951), all of whom have 
used psychoanalytic techniques, it appears that dreams from dif- 
ferent cultures frequently have strikingly similar latent contents. 
Typically, the analysis of non-Western dreams has revealed inces- 
tuous attachments, sibling rivalry, anxiety associated with castra- 
tion and maternal separation, cross-sex identification, and so forth. 
Roheim, a psychoanalyst and anthropologist, has used dreams con- 
sistently to illustrate Oedipal concerns in non-Western peoples. He 
has attempted to show, for instance, that among the Baiga such typ- 
ical Oedipal concerns as castration anxiety and hostility toward the 
father are prominent in fantasy, although the sexual behavior of 
both adults and children is subject to very few restrictions. 

Another man reported to Elwin the dream "I was in a rage, wrestling with 
my father; then a tiger knocked me down and killed me. I went below the earth, 
and there I turned into a tiny man only a foot high. A great snake saw me and 


said, "I am going to eat you." I said "Open your mouth," and in I went and 
came out the other end. At once I flew away; up to my own house." (Elwin 

The tiger is a representation of the father. The tiny man a foot high who goes 
into the earth is the dreamer's penis entering Mother Earth. The snake in this 
dream represents both father (phallus) and mother (devouring). Entering and 
flying are symbols of coitus. The latent dream wish is to kill the father (tiger) and 
have intercourse with the mother (Roheim 1946:507). 

Some of the findings about the universal characteristics of the 
latent content of dreams may be due to bias in the theory and meth- 
ods of dream interpretation, and especially to an overreliance on 
symbolic interpretation. This is not the case for all of these studies, 
however. For example, Clyde Kluckhohn's conclusions from his 
study of Navaho dreams are based not only on his informants' 
dreams, but on observations made after years of field work. The fol- 
lowing excerpt is taken from the analyses of a series of dreams from 
a five-year-old Navaho boy. 

Dream 2 

"We were in our hogan, and a wolf came, and he had long teeth, and he 
frightened us and Mamie ran to the bed, and I ran outside where my mother was 
and I hid behind her and she scared away the wolf. 

Yesterday I was playing in a deep arroyo. And above it my father was building 
on the adobe house. And I built some steps up the arroyo so I could climb out. The 
white dog and puppies came down into the arroyo. I got scared and couldn't find 
the steps. So I ran home. 

The dream itself is oedipal. The general pattern is already familiar from dreams 
of the other children. "Father threatens children. Mother protects us." Wolf in the 
dream and dog in the associations seem to be equated. The long teeth may repre- 
sent the penis, but they also recall the vagina dentata motif. Crawling out of the 
deep arroyo represents birth, and the whole of this part of the association suggests 
speculation about the father's part in the birth process. (Kluckhohn and Morgan 

In the same paper Kluckhohn states : 

I still believe that some of the cautions uttered by Boas and others on the pos- 
sible extravagances of interpretations in terms of universal symbols, completely 
or largely divorced from minute examination of cultural context, are sound. But 
the facts uncovered in my own field work and that of my collaborators have forced 
me to the conclusion that Freud and other psychoanalyists have depicted with 
astonishing correctness many central themes in motivational life which are uni- 
versal. The styles of expression of these themes and much of the manifest content 
are culturally determined, but the underlying psychologic drama transcends cul- 
tural differences. (Kluckhohn and Morgan 1951:120) 


Since these depth interpretations of dreams tend to find universal 
themes, is there any reason to include such investigations in the study 
of particular cultures? Roheim has argued strongly that dream in- 
terpretation can be extremely useful in uncovering the unconscious 
meaning of various cultural practices, such as initiation ceremonies 
and totemism. This is to be done by analyzing the context in 
which aspects of these practices occur in dreams ( Roheim 1932:21). 
For example, in the controversy over the supposed lack of knowl- 
edge of physical paternity of the Aranda, Roheim used the analyses 
of dreams concerning birth to indicate not only unconscious knowl- 
edge of the process of impregnation on the part of the Aranda, but 
also to illustrate that the official denial serves other personality needs, 
such as avoidance of rivalrous feelings toward the real father, and 
identification with the supernatural fathers, as well as disguised 
gratification of Oedipal wishes (Roheim 1938:359). 

It should be pointed out that it is the more unconscious and primi- 
tive dream contents which seem most universal rather than the par- 
ticular manner in which these impulses are expressed and defended 
against, and the specific reality situations which are associated with 
these impulses. For example, many of the dreams of Devereux's 
Plains Indian patient have the characteristics of moral maxims. In 
these dreams, significant figures often give the dreamer advice, help- 
ing him carry out his dream activities successfully. In other dreams 
the dreamer gives advice to himself, and to others. Here the mani- 
fest dream content shows a kind of moral life style typical of some 
individuals from the Plains culture area. This kind of manifest con- 
tent would be most unusual in Alor where theft and lying are more 
prevalent dream activities. 

In the last decade, there has been a shift of interest from the latent 
to the manifest content of dreams. The psychoanalytic concern 
with the functions of the ego and the potentialities of quantitative 
data treament have contributed to this change in interest. Within 
anthropology, Dorothy Eggan has pioneered in the study of the 
manifest contents of dreams. Mrs. Eggan has collected more than 
six hundred Hopi dreams from twenty informants in five Hopi 
villages, including over two hundred dreams with associations from 
a single informant who has been the subject of an extensive life 
study. Mrs. Eggan presents fifteen dreams from this informant, 
taken over a number of years, whose manifest contents demonstrate 
vividly some of the less conscious motivations in this man's life 
(Eggan 1949). The manifest content of this informant's dreams 


has also been subjected to a content analysis (Eggan 1952). The 
high ratio of the number of dreams which the dreamer felt were 
"bad" compared to the number experienced as "good," and the large 
number of dream elements dealing with security and support on 
the one hand, and elements of persecution and conflict on the other, 
portray the conflicts and unrest in the dreamer's personality. 

The shift in emphasis from latent to manifest content in the study 
of dreams, both Western and non-Western, has also been accom- 
panied by the use of more explicit theory and hypothesis in the con- 
struction of content categories (Hall 1956) . (C/. Eggan i960 for 
a further review of some of the current psychological research on 
manifest content in Western dreams.) Dittman and Moore (1957) 
have attempted to rate Navaho dreams for degree of emotional dis- 
turbance, comparing the dreams of members of the Peyote cult 
with nonmembers. They found that the members of the Peyote cult 
had somewhat more disturbed dreams, but only measured by the 
global dream ratings made by raters who had some knowledge of 
Navaho culture. 

Based on an earlier version of Schneider's analysis of the dream of 
the Yir-Yoront, a small Australian tribe, (Schneider 1941 ) , Walter 
Sears, in an undergraduate honors thesis, compared Navaho dreams 
with the dreams of the Yir-Yoront, and found that the Navaho 
have more threatening and terrifying dreams, fewer dreams in 
which aggression is directly expressed by the dreamer, fewer dreams 
with explicit sexual activities, and more dreams about white culture 
than the Yir-Yoront. Generally, it would seem that the Navaho are 
less free about the expression of impulses than the Yir-Yoront, and 
concomitantly find dreaming more unpleasant (Sears 1948). 

Another comparative study, by Griffith, Miyagi, and Tago, con- 
trasts the typical dreams of American and Japanese college students 
( 1958) . Griffith and his co-workers used the questionnaire method, 
requesting that the subjects check from a list of thirty-four typical 
dreams those which they could recall. It was found that there were 
great similarities between Japanese and Americans in the frequencies 
with which most typical dreams are recalled, with about as much 
agreement between the two cultures as there was between males 
and females within either culture. There were some small but sig- 
nificant diff^erences between the Japanese and Americans, however, 
the Japanese reporting more dreams of being attacked or pursued, of 
trying to do something again and again, of school, teachers, and 
studying, of being frozen with fright, of flying or soaring through 


the air, and of wild, violent beasts. The Americans report more 
dreams of arriving late, of missing trains, of being locked up, of 
loved persons being dead, of finding money, of being inappropri- 
ately dressed, of being nude, and of lunatics or insane people. These 
differences can be tentatively interpreted as indicating that Ameri- 
cans are more concerned with time, money, physical freedom and 
body shame than the Japanese, but less concerned with feelings of 
responsibility and projected aggression. 

An outline for an analysis of individual dreams, including both 
manifest and latent content, has been presented by Eric Erikson 
in a paper on the dream specimen in psychoanalysis (1954) . This 
outline breaks up the manifest content into verbal, sensory, spatial, 
temporal, somatic, interpersonal, and affective qualities, and the 
latent content into the sleep-disturbing stimulus, the day residue, 
acute life conflicts, repetitive conflicts, associated basic childhood 
conflict, impulses, and methods of defense. Using this outline, Erik- 
son reanalyses Freud's Irma dream (Freud 1900:106-120), illus- 
trating brilliantly not only the sexual and aggressive impulses woven 
into the dream, but also how the social and emotional conflicts at- 
tendant upon intellectual creativity were pictured by Freud in the 
dream, and solved within the dream by an individual ritual of iden- 

Adelson (i960) , using a modified form of Erikson's outline for 
the analysis of manifest content, has contrasted the dreams of col- 
lege girls who have highly rated literary creativity with the dreams 
of those who display little literary creativity. Creative girls tend 
to have dreams in which impossible events occur, often in an exotic 
setting, with many changes in these settings, while girls without 
much creative talent had dreams tied to the local, prosaic, and fa- 
miliar. Also, 20 per cent of the dreams of the creative girls were 
marked by the absence of the dreamer, while the noncreative girls 
always appeared in their own dreams. And finally, creative girls 
tended to have open and even flamboyant sexual activities occur in 
their dreams, contrasting sharply with the vague, timid, and sym- 
bolic dreams of the noncreative girls. Generally, it would seem that 
creative people, at least in dreams, can tolerate, and perhaps prefer, 
more incoherent, illogical, and directly expressive materials, and 
can treat these materials more impersonally. 

Recently, the study of dreams has received tremendous impetus 
from the psychological research on dreaming conducted by Dement, 
Kleitman, and others. They have found that an individual while 


dreaming moves his eyes much as he would if watching a play, and 
that such eye movements occur during light sleep, indicated by 
electroencephalographic records. If woken during rapid eye move- 
ment periods, an individual is able to report a dream approximately 
80 per cent of the time often in considerable detail. Even habitual 
nondreamers, who remember less than one dream a month, are able 
to report dreams on almost 50 per cent of the wakenings after rapid 
eye movements (Goodenough, Shapiro, Holden, and Steinschriber 
1959) . Eye movement periods, and presumably dreams, range from 
3 to 50 minutes in duration, averaging about 20 minutes, and tend 
to occur periodically throughout the night at intervals of 70 to 100 
minutes (Dement and Kleitman 1957) . The increase in both quan- 
tity and quality of dream reports made possible by this technique 
offers tremendous advantages in the study of dreams. Using this 
technique, Dement found that schizophrenics dream approximately 
the same amount of time as normals, but differ in their reports of 
dreams. Approximately half of the schizophrenic subjects fre- 
quently reported dreams of isolated, motionless, inanimate objects, 
apparently hanging in space. Dement rules out communication 
problems as the cause of this difference, but notes that although the 
schizophrenic subjects report motionless objects, their eyeballs were 
moving as if following moving objects. Dement concludes that this 
peculiar type of dream report is due to a "distorted schizophrenic 
concept of a more active visual experience" (1955:268). 

Dorothy Eggan reports, concerning a series of studies at Billings 
Hospital using Dement-Kleitman techniques, that the manifest 
content of dreams may show certain regularities over the course 
of an evening. 

Tresman, Rechtschaffen, OflFenkrantz, and Wolpert studied the patterning of 
dream content in two subjects over several nights of dreaming, while Offenkrantz 
and Rechtschaffen (1960a, 1960b) submitted the dream sequences of two addi- 
tional subjects to intensive clinical analysis. The results of these studies suggest 
the following: "While there is rarely a direct continuity of manifest content from 
dream to dream in the course of a night, a single emotional conflict, expressed 
in a variety of contents, may underlie all the dreams of an evening. There is a 
tendency for specific elements of manifest content to be repeated at similar times 
on different nights. The early dreams of a night tend to deal with events of the 
very recent past, often the experimental situation itself. Dreams of childhood 
scenes occur more often later in the night. (Eggan, personal communication) 

Findings based on two very different methods have been presented 
in this section. The first method, dream interpretation, uses symbols 
and associations to reconstruct the motivation which gave rise to 


the dream. The validity of this method obviously depends greatly on 
the skill and insight of the investigator. Using this method, very 
similar motives and conflicts have been found for peoples from dif- 
ferent cultures. 

The second method, content analysis, charts the frequency with 
which particular categories of dream events occur, making a com- 
parison of large samples of dreams possible. Using this method, dif- 
ferences in dreams have been found between such groups as the 
Navaho and the Yir-Yoront, creative and noncreative girls, schizo- 
phrenics and normals, and so forth. 

The difference in the type of findings reported for these two 
methods would seem to indicate that different levels of personality 
are being analyzed. It has been suggested that the individual dream 
interpretations have tended to refer to the more primitive and basic 
motivations similar in all cultures, while the content analyses have 
dealt more with the way in which impulses are expressed, defended 
against, and reintegrated, material which shows more individual 
and culturally distinctive patterning. 


Dreams, like other kinds of human behavior, can be expected to 
show some degree of cultural patterning. In dreams, however, con- 
scious self-control and external restraints, which serve as the two 
great agents of conformity with cultural norms, are almost com- 
pletely absent. Cultural patterning in dreams must come from deep 
within the individual rather than from conscious imitation of a 
cultural model, or the restrictions of cultural institutions. For this 
reason, cultural patterning in dreams seems especially relevant to 
an understanding of which aspects of cultural norms are most 
deeply internalized. 

One of the simplest and most direct ways in which culture might 
be expected to affect dreams is in manifest content. Certainly, peo- 
ples who have never seen automobiles are not likely to dream of 
them. However, there is evidence that dreams do not give a faith- 
ful point-by-point representation of the sector of culture experi- 
enced and manipulated by the individual in waking life, but instead 
give a selective, edited picture of the individual's cultural world. 

First, some dreams seem almost completely bare of cultural items 
of any sort. Often these dreams are symbolic dreams of flying, body 
destruction, landscapes, animals, and so forth. Perhaps the cultural 
bareness of these dreams is due to the difficulty in translating dream 


images into words, and then retranslating into the ethnographer's 
language. I would guess that about one fifth of the dreams I have 
examined from non-Western cultures lack any culturally distinc- 
tive materials in manifest content, although this seems to vary by 

Second, certain areas of cultural life are overrepresented in the 
manifest content of dreams, while other areas may be considerably 
underrepresented. Within the United States, Calvin Hall finds that: 

Dreams contain few ideas of a political or economic nature. They have little 
or nothing to say about current events in the world of affairs. I was collecting 
dreams daily from students during the last days of the war with Japan when the 
first atomic bomb was exploded, yet this dramatic event did not register in a 
single dream. Presidential elections, declarations of war, the diplomatic struggles 
of great powers, major athletic contests, all of the happenings that appear in news- 
papers and become the major topics of conversation among people are pretty 
largely ignored in dreams. 

What then is there left to dream about? There is the whole world of the per- 
sonal, the intimate, the emotional and the conflictful, and it is this world of ideas 
out of which dreams are formed. (1953:1 1— 12) 

Emotionally, the content of dreams seems to contain more nega- 
tive feelings than waking life. In a content analysis of a large sample 
of Western dreams, Hall found that 40 per cent of the emotions 
displayed in dreams can be characterized as apprehension, 18 per 
cent as anger, and 6 per cent as sadness. Another 1 8 per cent of the 
emotions are characterized as neutral excitement and surprise, while 
only 1 8 per cent are characterized as happiness. In this same sample 
almost half of the dream persons were strangers to the dreamer, 
while about 20 per cent were family, of which 34 per cent were 
mother, 27 per cent father, 14 per cent brother and 12 per cent 
sister (Hall 1951) . 

The manifest content of dreams may also reflect the sex of the 
dreamer. Hall finds that men in our culture dream about males 
twice as frequently as about females, while women dream equally 
about both ( Hall 1 9 5 1 ) . In Lee's study of Zulu dreams ( 1 9 5 8 ) , an 
unusual degree of difference between the manifest contents of the 
dreams of men and the dreams of women was found; the women 
dream more of babies and children, the men dream more of fighting 
and cattle. This difference reflects the traditional division of labor, 
although at the time of the study, the traditional separateness of 
men's and women's activities had broken down. Lee's hypothesis, 
that the content of dreams is laid down in the early years of life. 


oflfers an interesting avenue of exploration, which might account 
for the lack of political and economic activities noted by Hall. 

Devereux offers a similar hypothesis about the relation between 
the dream content and childhood experience of a Plains Indian in 
psychotherapy. The items of aboriginal culture which appeared in 
this patient's dreams were those which "reflected most clearly both 
the highest traditional values of Wolf culture (pseudonym for the 
patient's culture) , and the least rational parts thereof: i.e., medicine 
bundles, magic and the like" (1951:100). These aboriginal mate- 
rials began to appear with greater frequency in the patient's dreams 
when he began to analyze his own past, and dreams with many ab- 
original items were often the most significant and revealing. 

Devereux speculates that the small amount of manifest content 
taken from the immediate present in this patient's dreams may be 
due to the fact that these dreams reflected life-long defense mechan- 
isms, laid down in childhood, and also to the fact that Plains Indian 
children were often brought up by their grandparents, who embody 
the more traditional culture (1951:88). 

Holmberg, in his study of the Siriono, found the manifest con- 
tent of dreams to be related to one of the central features of Siriono 
life. The Siriono are a hunting and gathering people of the interior 
Amazon, who are often if not always hungry and who spend much 
of their time in a grim search for food. Holmberg found that more 
than half of a sample of fifty dreams were concerned with eating 
food, hunting game, and collecting edible products from the forest. 
One of the most common dreams is that a relative out hunting has 
had luck and is returning with game for the dreamer (Holmberg 
1950:91). He found that "one of the striking things about food 
dreams is that they seem to occur just about as often when a person 
is not hungry as when he is hungry" (Holmberg 1950:91). This 
would lead one to speculate that food has come to symbolize a num- 
ber of things for the Siriono besides its hunger-reducing properties. 

So far some of the ways in which dreams tend to give a selective 
and edited picture of the dreamer's culture have been described. 
Schneider and Sharp, in a thorough and systematic monograph, 
have investigated the relation between Yir-Yoront dreams and cul- 
ture. The dreams were collected by R. L. Sharp, and analyzed by 
D. Schneider (in manuscript) . Schneider begins with the assump- 
tion that dreams portray the dreamer's view of the world, or his 
"definition of the situation," and that culture, as a system of norms, 
afl^ects but is not identical with this definition of the situation. 


In order to investigate the relation between Yir-Yoront dreams 
and culture Schneider has analyzed the manifest content of 149 
dreams taken from 51 subjects, 43 men and 8 women. Four kinds 
of dream situations were studied; dreams involving sex, aggression, 
death, and contact with white culture. Certain striking regularities 
in these areas were uncovered. Nineteen dreams containing explicit 
material on sexual intercourse were found, all from men. The part- 
ner in these dreams is in a little more than half of the cases from the 
approved classificatory kinship class (mother's brother's daughter) , 
although in only one case is the sex partner actually a wife. 

Perhaps the most interesting finding from a review of the dreams of sexual 
intercourse is that when the sex partner is of a prohibited degree of relationship, 
and where no adjustment to this fact has been made in waking Ufe, the 
men picture ( i ) a specific interruption before or during the act of intercourse 
which occurs as (a) an organic defect of the woman's sexual organs or (b) an 
overt, verbal rejection of the male dreamer's advances which have little deterrent 
effect in the dream. (2) The magnitude of the interruption correlates with the 
strength of the prohibition on sexual relations. Intercourse with FaSiDa never 
gets started; with the SiDa, the act is completed but with difficulty; with the 
SiDaDa there is merely verbal rejection of the man by the woman. (Chapter 

Another interesting finding involves the expression of aggression. 
In Yir-Yoront dreams both mother's brother and elder brother are 
frequent aggressors against the dreamer. This is quite different from 
the actual situation, in which a man gives gifts and shows respect 
towards his mother's brother, and treats his older brother with 
deference. Dreams involving death also show some surprising pat- 
terning. While there is no cultural belief in resurrection, in most 
of the dreams of death in which the dreamer himself dies, the 
dreamer then "stands alive" or is resurrected. However, in dreams 
in which someone else dies, the corpse most frequently remains dead. 

These findings raise some interesting questions about the relation 
between any fantasy product, such as dreams, and the actual ex- 
periences of the individual. Certainly most fantasy, including 
dreams, contains something of a "reflection" of the individual's ex- 
periences and his "definition of the situation." Usually this "re- 
flected" material is selected and edited according to the particular 
interests and concerns of the person. For example, it has already been 
mentioned that personal and intimate materials are more likely to 
appear in American dreams than public and political matters. 

Selection and editing of fantasy, however, results in only mild 
distortions of the individual's actual experience. Sometimes the di-^- 


tortion is more drastic, as in obvious cases of wish fulfillment. 
Schneider and Sharp consider the fact that the sex partner in Yir- 
Yoront dreams is almost always some one other than a wife to be the 
result of wish fulfillment, and, in a sense, still a part of the indi- 
vidual's definition of the situation. 

Projection is a still more drastic kind of distortion. Certainly the 
Yir-Yoront tendency to picture mother's brother and elder brother 
as hostile and aggressive, when the shoe is on the other foot, would 
seem to fit neatly the definition of projection. Other dream mate- 
rials may also involve projection, but are less discernible because the 
individual's actual experiences are less clearly known. For example, 
it may be that the dreamer's portrayal of the woman as the source 
of interruption of intercourse is pure projection, or this may be an 
accurate portrayal of what actually happens. If it is projection, this 
would make some sense out of such bizarre items as the woman's 
clitoris falling off in one of these dreams, and the other images of 
the woman having damaged genitals. It would then be really the 
man's genitals which would become injured, or which he fears would 
become injured if intercourse with a forbidden woman were to take 
place. Here the dreamer's actual "definition of situation" is reversed, 
although the anxiety is still apparent as sexual in origin. 

An even more elaborate kind of distortion occurs in instances of 
symbolization. For example, it may be that the resurrection dreams 
of the Yir-Yoront are symbolic dreams of repeated sexual inter- 
course, in which the penis dies and is then born again. The men of 
the Yir-Yoront "in waking life, talk as if a single act of intercourse 
was more unusual than four, five, or six," but in overt sex dreams 
rarely have more than one act of intercourse. Perhaps the same anx- 
iety that gives rise to this kind of bragging also motivates the "stand 
alive" dreams. The following dream of a mature man may be a case 
in point, and gives something of the flavor of Yir-Yoront dreams 
in general. 

I'm making a forked support for the corpse at Olwin-an. It is for Spear's sister 
(dead, unknown) . I saw Yaltide's vagina. Her legs were far apart. A mob from the 
north (Yir Ma'as and others) speared me. I lay down alongside the corpse. I 
was full of spears. The North people cut me up. They took my bones out. They 
cut me up like a wallaby. They ate my liver and flesh after cooking it. I came 
alive again. I had healed up but had no bones, which had been smashed up and 
the marrow eaten. My brains, bones, etc. were all eaten. Wil (also was eaten). I 
rolled up belongings and left. I went along and died. I was buried. I heard people 
keening for me. Women were jabbing sticks in their vaginas so that blood would 
run out; they were sorry. Blood running down their legs, vaginas. I came alive 


again. Stretched arms and legs and back. I went off hunting. I killed two goannas, 
cooked them and woke up. 

Inf.: Parkaia perhaps sent dream. My mother, who is Spear's sister (dead, un- 
known), was dead in the dream. Yaltelde, my sister, was simply mourning the 
corpse. I dreamed this last night, (dream 34) 

If this hypothesis is correct, it would help explain why "dreams 
of death are noticeably lacking in intense affect," and why resurrec- 
tion occurs in dreams but not as an item of cultural belief. In any 
case, the relation between the culture, the individual's experience, 
and dreams of "standing alive" is not a simple one. 

To summarize so far, it seems that there is no simple relation be- 
tween culture and the manifest content of dreams. This appears to 
be because a dream is not exclusively a cognitive act, in which things 
once perceived are reshuffled and reviewed in the mind's eye. Instead, 
the dream is a selective, edited, and sometimes highly distorted ver- 
sion of the individual's experience. This selectivity and distortion 
is generally considered to be an effect of motivation, as well as the 
type of special mental process involved in dreaming. The various 
examples of selectivity in dream content mentioned above, such as 
the frequent reference to food in the dreams of the Siriono, the sex 
differences in the dreams of the Zulu, an acculturated Plains In- 
dian's tendency to dream about nonrational aspects of his aboriginal 
culture, and Hall's finding that American dreamers dream about 
the personal and intimate rather than the political and economic, 
would, therefore, be held to be due to the particular needs of indi- 
viduals in these societies. More dramatic distortions seem to be 
due to conflict. The Yir-Yoront projection of hostility onto the 
mother's brother may represent such a conflict, perhaps in this case 
between aggressive feelings and anxiety about retaliation. 

The effect of culture on dreams may be seen more directly in the 
"culture pattern dream." (Lincoln 1935:189). These dreams, 
which are specified and sanctioned by the culture, and which usu- 
ally involve supernaturals or supernatural manifestations, are often 
considered visions (Lincoln 1935:189). The Crow, for example, 
gave great importance to culture pattern dreams, and success in 
life was considered to depend upon these visions. Lowie remarks 
that he never succeeded in securing a detailed narrative of an ordi- 
nary dream, because his informants would report only visions 
(1922:342) . Typically, culture pattern dreams of this type involve 
a preparatory phase of fasting, isolation and self-mutilation, fol- 
lowed by a hallucinatory experience, in which a spirit helper, usually 


in human guise, adopts the dreamer as his child, and gives him spe- 
cific instructions in the use of a supernatural power. 

Although it might seem likely that individuals would falsify 
such experiences in order to gain honor and riches, Lowie reports 
that this was not the case. In fact, some people were never success- 
ful in obtaining a vision, and others, who thought they had received 
a true revelation, later became convinced through testing their sup- 
posedly acquired powers that they had been deceived by their vision 

In those cases in which an individual believes that he has had a 
culture pattern dream, the degree to which the content of the dream 
has been affected by secondary elaboration, in which the dreamer 
unwittingly assimilates the dream experience to a previous cultural 
model, remains problematic. Sometimes such a process of secondary 
elaboration can be seen quite clearly. Erika Bourguigon notes that 
in Haiti a dream may be recounted as if a particular supernatural 
had appeared in it, although more detailed questioning would reveal 
that only an ordinary person with certain characteristics which 
might indicate a disguised supernatural had been seen in the dream. 
Her conclusion, based on Haitian materials, is probably represent- 
ative for other societies in which culture pattern dreaming occurs. 

While it is difficult to see to what extent dreams themselves may be culturally 
patterned, the cultural dogma of the dreams as appearance of the gods interacts 
with the dream content in such a way that an interpreted version of the dream 
seems to be experienced by the dreamer. (E. Bourguigon 1954:268) 

The effect of acculturation on culture pattern dreaming has 
been discussed by Radin and King. Radin presents some evidence 
that as a result of acculturation, the Ottawa and Ojibwa stopped 
having culture pattern dreams and began to have dreams concerned 
only with personal problems (Radin 1936). King documents the 
opposite case, in which an acculturated Mountain Maidu Indian 
(whose biological father was white) had a series of culture pattern 
dreams which incorporated elements of Western culture (King 
1943) . In this series of dreams the dreamer was able to defeat the 
magical attacks of malicious shamans by using both Indian and 
white kinds of magical power. King finds that the remarkably good 
adjustment of this man to Western culture is shown in these dreams, 
and also speculates that dreams might be fruitfully used to study 
psychological adjustment in acculturation. 

Cultural beliefs and theories about dreams also appear to affect 


the content of dreams and emotional reactions to dreams. One often 
quoted example of the effect of dream theories is the difference be- 
tween the Tikopia and the Trobriand Islanders in their emotional 
reactions to incest dreams (Firth 1934). The Tikopia believe that 
incest dreams are inspired by malignant spirits who may imperson- 
ate relatives and seduce the dreamer. The Trobriand Islanders, on 
the other hand, believe more in the reality of their dreams, and react 
with shame and guilt to incest dreams. While the Tikopia do not re- 
act with shame and guilt, they nevertheless do not completely escape 
the consequences of such dreams. For the Tikopia, sexual intercourse 
in a dream is sexual intercourse with a spirit, and intercourse with 
spirits results in loss of vitality and illness. In general, Tikopia dream 
theory demands taboo on sexual intercourse as a goal of the dream, 
while Trobriand dream theory involves a taboo only on certain 
sexual objects, a difference which may correspond to personality 
features characteristic of these two societies. 

A somewhat more subtle effect of dream theories on dreams has 
been noted by Devereux, who points out that where dreams are 
given certain kinds of objective reality, the dreams of individuals ap- 
pear to be more egosyntonic, and in such cultures dream events tend 
to be more similar to real life events, and also to be more useful to the 
individual, who may use his dreams to plan new activities, and to 
attempt to integrate old and painful experiences by reworking them 
successfully (1951:87). 

Another possible effect of dream theories has been explored by 
Hallowell, who finds that where dreams are considered to be actual 
experiences of the self, as among the Objiwa, that the self may be 
conceived of and experienced as capable of dream-like activities, 
such as physical metamorphosis, separation from the body, and the 
ability to shift back and forth in time. As a result of such a self- 
conception, and the integration of dreams with waking experiences, 
the "behavioral environment" or "habitat" of the individual may 
come to have radically different qualities than the "physical en- 
vironment" (1955:172-182). 

The findings concerning culture pattern dreams suggests that, 
while it is possible for some individuals to dream as required, this 
is not an easy task, and for some persons even impossible. Some of the 
implications of this situation will be discussed below. 

With respect to the ways in which cultural cognitive structures 
affect dreams, the findings of Firth, Devereux, and Hallowell sug- 
gest that native dream theories play a part in taming fantasy, mak- 


ing it more like waking life, and, reciprocally, in making waking life 
more like dream fantasy. 

The Cultural Uses of Dreams 

In the ethnographic literature a wide range of beliefs and prac- 
tices concerning dreams has been reported. These beliefs and prac- 
tices enter into many different aspects of culture. One important set 
of culture traits relates dreams to the religious system, and includes 
the use of dreams to contact and gain power from supernaturals, 
as well as the more common beliefs that the soul wanders during 
dreams, meets other souls, and is responsible for its actions. Another 
set of traits concerns the use of dreams in the social system, in which 
there may be formal or informal statuses and roles involving dreams, 
such as dream interpreters, or shamanistic dream performances, and 
roles which can only be assumed if the proper dream is dreamed. An 
almost universal set of traits involves the use of dreams to predict 
the future. The last major group of traits involves emotional ca- 
tharsis through ritualized methods of reacting to dream experi- 
ences, in which the effect of a bad dream may be dispelled or a good 
dream made to come true by a more or less elaborate ritual, such 
as not telling the dream, or acting out the dream commands, or 
making a sacrifice. 

These traits are not cultural monads, but have functional rela- 
tions with other phenomena, cultural, social, and individual. Two 
examples of such relations, concerning "primitive dream psycho- 
therapy" and "unconscious role acceptance" have been discussed in 
the anthropological literature. 

"Unconscious role acceptance" becomes a factor in a social sys- 
tem when culture pattern dreams are used to determine which roles 
an individual will assume. The dreamer may either be obligated to 
assume a particular role because he has had a certain type of dream, 
as, for example, among the Sioux, where dreams of the moon, or a 
hermaphroditic buffalo, require the individual to become a her- 
dache, or the dreamer may be required to dream a particular culture 
pattern dream before he is allowed to assume a certain role, as among 
the Pukapuka, where qualifications for priesthood require that a 
man have dream contact with supernatural powers during the ini- 
tiation period. 

Since dreams are not under direct conscious control, the use of 
culture pattern dreams to determine role taking brings factors of 
"unconscious choice" into consideration. A young man who is re- 


quired to have a vision and obtain a spirit helper before he may have 
all the responsibilities and privileges of the adult role may con- 
sciously want to assume an adult role, but if on a less conscious level 
he feels he is not ready to become a man, dreaming the required 
dream would probably be an impossibility, both because of uncon- 
scious sabotage, and because typically the content of the culture 
pattern dream in these cases is psychologically sound, symbolizing 
accurately the resolution of dependency conflicts. Also, where an 
individual is forced into a deviant role because of his dreams, not 
only are unconscious factors taken into account, but a culturally 
legitimate excuse is given for such deviancy. Erikson, in his discus- 
sion of the Sioux, states: 

A homogenous culture such as that of the Sioux, then, deals with its deviants 
by finding them a secondary role, as clown, prostitute, or artist, without, how- 
ever, freeing them entirely from the ridicule and horror which the vast majority 
must maintain in order to suppress in themselves what the deviant represents. 
However, the horror remains directed against the power of the spirits which have 
intruded themselves upon the deviant individual's dreams. It does not turn against 
the stricken individual himself. In this way, primitive cultures accept the power 
of the unconscious. As psychopathologists, we must admire the way in which these 
"primitive" systems managed to maintain elastic mastery in a matter where more 
sophisticated systems have failed. (1950:137) 

Another example of the use of dreams to manage psychological 
problems can be found in primitive psychotherapy, discussed by 
Wallace (1958) , Devereux ( 1 951) , Kilton Stewart (1954, 1951), 
and Toff elmier and Luomala (1936). Generally, such therapy seems 
to consist of a cultural recognition that dreams reveal hidden wishes 
and conflicts, and a culturally prescribed method of dealing with 
those wishes and conflicts. The most common method of handling 
such wishes and conflicts seems to be to fulfill or act out the wish, 
once it is revealed. Anthony Wallace presents an impressive example 
of this method in his study of Iroquois dream theory. 

Intuitively, the Iroquois had achieved a great deal of psychological sophistica- 
tion. They recognized conscious and unconscious parts of the mind. They knew the 
great force of unconscious desires, and were aware that the frustration of these 
desires could cause mental and physical ("psychosomatic") illness. They under- 
stood that these desires were expressed in symbolic form by dreams, but that the 
individual could not always properly interpret these dreams himself. They had 
noted the distinction between the manifest and latent content of dreams, and 
employed what sounds like the technique of free association to uncover the latent 
meaning. And they considered that the best method for the relief of psychic and 
psychosomatic distress was to give the frustrated desire satisfaction, either directly 
or symbolically. (1958:237—238) 


Among the Senoi, impulses revealed in dreams are evidently han- 
dled in an unusually sociable fashion, so that if a man dreamed he 
was attacked by another, he would attempt to settle the differences 
between them through discussion and mediation (K. Stewart 195 1 ) • 
The Navaho, on the other hand, use dreams not to reveal wishes, but 
to indicate proper curing rituals. Lincoln suggests that the Navaho 
curing ceremonies prescribed on the basis of the content of dreams 
have symbols similar to those of the diagnostic dreams, and that the 
particular curing ceremony is effective because it resolves sym- 
bolically the unconscious conflict in the dream. For example: 

Dreams of death, that is, of one's own death, or the death of neighbors and 
relatives, also dreams that your teeth have fallen out require the Hozhonju or 
Chant of the Restoration of the Family. 

Sttggestion. Death dreams are generally death wishes, and the symbol of losing 
a tooth as often meaning castration anxiety because of death wishes is widespread. 
(Here again occurs the association of loss of a tooth, death of a relative as in the 
universal type dreams.) The Hozhonji is to restore the family, that is to protect 
it from death wishes towards the parents. (Lincoln 1935:180) 

It has been suggested by Stewart that a therapeutic psychological 
effect may be obtained if the symbolic forms which emerge in trance 
and dream are taken as objective dangers, and group support is 
given to mastering these symbolic dangers. Stewart presents a vivid 
if journalistic account of the psychotherapeutic methods of the 
Phillipine Negritos. A group of shamans co-operate in placing the 
patient in trance, and then encourage the patient to meet and over- 
come the spirit that has caused the patient's illness. This spirit, which 
has been attacking the patient in his dreams, is made to give the 
patient a song, and to become the patient's spirit helper. Stewart 
comments that this method seems effective in curing chronic physi- 
cal ailments, such as skin irritations, headache, and recurrent fever, 
which probably have at least a partial psychosomatic origin. In this 
form of therapy, conflicts are externalized as spirits, and group 
support is given to overcoming their symbolic representations. Also, 
a spirit, once faced and overcome, is made to work for the person, 
and a public ritual is used to displace previous anxiety (Stewart 


A technique of dream therapy has been reported for the Diegueno 

Indians of southern California which seems to be similar to Western 

psychotherapy in its management of dreams. This technique is used 

to treat persons who appear to be afflicted with obsessive sexual 


fantasies. There are two recognized forms of this type of illness. 
The first, which is less serious, is characterized by symptoms of ex- 
cessive dreaming, laziness, and social withdrawal. The second form 
of this malady is considered to be an advanced form of the first, 
and appears to be an actual psychosis, characterized by persistent 
hallucinations of a spirit lover, a supernatural bullet hawk which 
takes human form as a person of either sex. Persons afflicted with 
this hallucination are called "spouses of that bird." 

To treat these maladies, a dream shaman is sought. The shaman 
attempts to get the patient to talk about his dreams and sexual life, 
actual and imaginary. The shaman begins by asserting that he al- 
ready knows all the patient's dreams, so that there is no use in trying 
to conceal anything. A mild type of hypnotic trance may be used 
to encourage the patient's talking, except in the more severely 
psychotic cases, which do not respond to this kind of treatment. 
Along with discussion of the patient's sexual life and fantasies, the 
shaman also prescribes blood letting and special nourishing foods. 
For the unwed, marriage is recommended, apparently to help the 
patient shift from substitute gratification in fantasy to real life 
situations (Toffelmier and Loumala 1936) . The technique of ther- 
apy in this example is in many ways unusual. The technique of dis- 
cussing with the patient his fantasies, including dreams, rather than 
permitting the patient to enact his fantasies, or to create a ritual 
defense against them, is particularly striking. It is not surprising 
that in this culture shamans are selected because of their stable 
(rather than unstable) personalities. 

To summarize the material which has been treated so far in this 
section, the distinctions between content, structure, function, and 
process in culture may provide a useful framework (Hsu 1959). 
The cultural uses of dreams may be considered to be a type of cul- 
ture content, having relations with the structural, functional, and 
procedural aspects of culture and society. Dreams may affect the 
structure of a society in becoming the subject matter of formal and 
informal roles, such as that of the dream interpreter, or in becoming 
a prerequisite for the ascription and achievement of roles, bringing 
factors of unconscious choice into the process of role allocation, as 
well as offering justification for the choice of deviant roles. Dreams 
may also function to help the individual maintain psychic equi- 
librium, serving as an important part of non- Western and Western 


Dream Usages and Their Correlates 

The next part of this section will present the results of a cross- 
cultural study of the conditions which affect the cultural uses of 
dreams. In this study I have attempted to find out why some socie- 
ties have extensive uses for dreams, while other societies do not. 
On the basis of case history materials reported in the ethnographic 
literature, it seemed to me that anxiety about being alone and on 
one's own often gives rise to a strong preoccupation with dreams 
and fantasy. If this were true, then societies in which individuals 
frequently experience anxiety concerning isolation and self-reliance 
would be likely to place an especially strong cultural emphasis on 
dreams. I therefore attempted to specify the social conditions which 
would be most likely to subject individuals to this type of anxiety, 
so that it would be possible to predict the degree of emphasis placed 
on dreams in any given culture from these conditions. 

Field workers interested in culture and personality have presented 
several examples of the effect of social isolation and the effect of cul- 
tural roles which demand independent and self-reliant action. Mar- 
garet Mead recounts the story of an orphaned Manus boy who felt 
isolated and unloved, and who, unlike the other Manus children, was 
preoccupied with fantasies about a guardian spirit which he took 
to be his own father (Mead 1932:183) . A similar case has been re- 
ported by Dorothy Eggan, in a study of mythic materials in dreams 
(1955). One of her Hopi informants, who also felt isolated and 
abandoned, also turned inward to fantasy about a supernatural 
helper. Dorothy Eggan comments: 

Benedict has pointed out that although the Pueblo area is surrounded by the 
concept of a power-giving or protecting Guardian Spirit, such a concept has not 
been standardized in the Pueblo groups because they are dominated by the "neces- 
sity of the group ceremonial approach not that of individual experience" (Bene- 
dict 1923:36). But in Sam we find a man who, because of personal problems, 
although believing firmly in the "group approach," was frequently made to feel 
less a part of the community than he needed to feel. Consequently he has elaborated 
the concept of dtimalaitaka (guide or guardian spirit) , which is found among the 
Hopi, but which is generally rather vague and unstressed, into an ever present and 
active spirit who comes to him in dreams, takes him to witches' meetings and on 
treasure hunts, gives him strength, wisdom and advice, rescues him from danger- 
ous situations, and always assures him that he is on the right road and that his 
enemies are wrong. (1955:448) 

Wallace, in his study of Iroquois dream theory, also concludes 
that anxiety about independence is related to this kind of extensive 
use of dreams: 


. . . the typical Iroquois male, who In his daily life was a brave, generous, active, 
and independent spirit, nevertheless cherished some strong, if unconscious, wishes 
to be passive, to beg, to be cared for. This unallowable tendency, so threatening 
to a man's sense of self-esteem, could not appear easily even in a dream; when it 
did, it was either experienced as an intolerably painful episode of torture, or was 
put in terms of a meeting with a supernatural protector. However, the Iroquois 
themselves unwittingly make the translation: an active manifest dream is ful- 
filled by a passive receiving action. The arrangement of the dream guessing rite 
raises this dependency to an exquisite degree: the dreamer cannot even ask for his 
wish; like a baby, he must content himself with cryptic signs and symbols until 
someone guesses what he wants and gives it to him. (1958:247). 

These reports indicate that anxiety about being isolated and on 
one's own may give rise to preoccupation with dreams and fantasy, 
especially fantasy about magical helpers. The content of such fan- 
tasy seems to serve as a denial of the individual's actual isolation and 
helplessness, thereby partially relieving these anxieties. 

In order to measure the degree of cultural preoccupation with 
dreams, the following traits involving dreams were coded for a sam- 
ple of sixty-three societies taken from the Human Relations Area 
Files. No society was selected unless at least a paragraph on dreams 
could be found in the literature, and no more than two societies have 
been taken from any one culture area, using Murdock's World 
Ethnographic Sample (1957) : 

a) Supernaturals appear in dreams and give important powers, aid, ritual, and 

b) Religious experts (priests, shamans) expected to use their own dreams in 
performance of their role (e.g., curing, divination) . 

c) Culture pattern dreams required before some roles may be assumed. 

d) Dreams induced by special techniques (e.g., fasting, drugs, sleeping alone, 
etc.) . 

e) Formal or informal role of dream interpreter. 

/) Undoing ritual after some dreams (e.g., sacrifice, avoidance). 
g) Supernaturals appear in dreams and harm or foreshadow harm to the 

These particular traits were selected because they are neither uni- 
versal nor extremely rare, and because they cover a wide range of 
types of uses of dreams. I had hoped that all of these traits would be 
positively correlated with each other; however, this proved not to 
be the case. Only four of these traits showed high significant corre- 
lations with each other: traits a, b, c, and d. The other three traits 
were uncorrected with each other, and with these four. 

If all seven traits had been strongly intercorrelated, it would have 
been reasonable to assume that there is a general factor of preoccu- 


pation with dreams. The findings seem to indicate, however, that 
rather than a general factor of preoccupation with dreams, there is 
a more hmited complex centered about the use of dreams to seek 
and control supernatural powers. Traits a, b, and d involve this 
seeking and controlling of supernatural power quite directly. Trait 
c, involving culture pattern dreams which are required before cer- 
tain roles may be assumed, is less directly related to seeking super- 
natural aid. However, it seems that such culture pattern dreams 
often consist of a visitation by a magical helper, who teaches the 
aspiring shaman or warrior important supernatural techniques. The 
other dream traits, involving dream interpretation, undoing rituals 
and possible supernatural harm, are unrelated to this complex, and 
have not been used in measuring this type of cultural preoccupation 
with dreams. 

In view of these findings, the original hypothesis has been modi- 
fied to state that anxiety about being alone and on one's own gives 
rise to the use of dreams to seek and control supernatural powers. 
The extent of this use of dreams has been measured by the number 
of traits a, b, c, and d reported present for each society. The median 
number of traits reported present for this cross-cultural sample is 
one. Societies with none of these four traits fall below the median, 
and are considered low on the use of dreams to seek and control su- 
pernatural powers. Societies with one or more traits reported pres- 
ent are considered high on this use of dreams. 

The first condition specified as a possible cause of anxiety about 
isolation and independence involves residence at marriage. If, at 
marriage, a son or daughter moves far away from his or her parents, 
the loss of parental support should give rise to anxiety about being 
isolated and on one's own. In order to test this hypothesis, estimates 
of the distances that sons and daughters most usually move at mar- 
riage for each society have been taken from a cross-cultural study 
of residence by Whiting and D'Andrade (1959). Table i presents 
the association between the typical distances for parents and mar- 
ried son and the use of dreams to seek and control supernatural 
powers. The data in this table indicate that the further the son 
typically moves away from his parents, the more likely a society is 
to use dreams to seek and control supernatural powers. The degree 
of association is fairly strong, and significant at the .01 level. No 
table has been presented for the relation of distance between parents 
and married daughter and use of dreams because it was found that 
there is no association between these two measures. Apparently, 




Relation of Most Typical Distance Between Married Son and Parents to 

Use of Dreams to Seek and Control Supernatural Powers 

The societies are grouped in columns on the basis of distance between married son and parents 

and in descending degree of extensiveness of use of dreams to seek and control supernatural 

powers. The letters in parentheses after each society designate the traits reported present. (See 

page 321 for definition of traits.) 

Son Resides 
in Parents' 

Son Resides 
in Same Village 
or Local Group 

Son Resides in 
Different Village 
or Local Group 

Papago {a,b,c,d) 
Kapauku {a,b) 
Ifugao (c) 

Comanche {a,b,c,d) 
Semang {a,b,c,d) 
Pukapuka {a,b,c) 
Chukchee {a,b) 
Rwala {b,d) 
Araucanians {b) 
Azande {d) 
Fang (a) 
Nyakusa {b) 
Wolof {b) 

Crow {a,b,c,d) 
Iroquois (a,b,c,d) 
Jivaro (a,b,c,d) 
Naskapl (a,b,c,d) 
Ojibwa {a,b,c,d) 
Omaha (a,b,c,d) 
Paiute {a,b,c,d) 
Andamans {a,b,c) 
Copper Eskimo {a,b,c) 
Cuna (a,b,c) 
Kaska {a,c,d) 
Lapps {a,b) 
Yaruro {b,d) 
Bemba {b) 
Mundurucu (a) 
Trobriands (r) 
Yakut ((/) 

Bhil (-) 
Iban (-) 
Lepcha (-) 
Mataco (-) 
Nama (-) 
Samoa (— ) 
Siriono (— ) 
Tupinamba (— ) 

Ashanti (-) 
Aymara (-) 
Ifaluk (-) 
Kurtatchi (-) 
Marquesas (— ) 
MinChia (— ) 
Mossi (— ) 
Riffians (-) 
Somali (-) 
Tallensi (-) 
Tanala (— ) 
Thai (-) 
Tiv (-) 

Tubatulabal (— ) 
Yoruba (-) 

Burmese (— ) 
Callinago (— ) 
Ganda (-) 
Karen (— ) 

anxiety suffered by women does not affect this use of dreams. Per- 
haps this is because reHgion is more frequently a man's affair, or 
perhaps because women may turn to their spouses in order to reheve 
the anxiety of loss of parental support in a way that men may not. 
In order to check on these findings, Murdock's residence and 
family classification has been used. From the findings presented 
above, nonpatrilocal societies should have more uses of dreams to 



seek and control supernatural powers than patrilocal societies, and 
independent families should have more uses for dreams than ex- 
tended families. Both these conditions have effects in the predicted 
direction and are statistically significant when considered together. 


Relation of Subsistence Economy to Use of Dreams to Seek and Control 
Supernatural Powers 
The societies are grouped in columns on the basis of economy in descending degree of extensive- 
iiess of use of dreams to seek and control supernatural pwwers. The letters in parentheses after 
I'ach society designate the traits reported present. (See page 321 for definition of traits.) 



Hunting, Fishing, and 



Animal Husbandry 







Comanche {a,b,c,d) 

Crow (a,b,c,d) 

Naskapi {a,b,c,d) 

Ojibwa {a,b,c,d) 

Omaha (a,b,c,d) 

Paiute {a,b,c,d) 

Semang {a,b,c,d) 

Andaman {a,b,c) 

Iroquois {a,b,c,d) 

Copper Eskimo {a,b,c) 

Jivaro (a,b,c,d) 

Kaska {a,c,d) 

Papago {a,b,c,d) 

Pukapuka (a,b,c) 

Cuna {a,b,c) 

*Chukchee (a,b) 

Carib {a,b) 

''Lapps ia,b) 

Azande {d) 

*Rwala {b,d) 

Chagga {b,c) 

Bemba {b) 

Wishram (a,d) 

Kapauku {a,b) 

Fang {a) 

Yaruro {b,d) 

Araucanians {b) 

Ifugao (r) 

Caingang (d) 

Nyakusa {b) 

Mundurucu {a) 

Tlingit (^) 

Wolof {b) 

Trobriands (c) 

* Yakut (a) 

Aymara (-) 

Ashanti (-) 

Callinago (-) 

Bhil (-) 

Ifaluk (-) 

Mataco (-) 

Burmese (— ) 

Kurtatchi (-) 

=^Nama (-) 

Ganda (-) 

Marquesas (-) 

Siriono (-) 

Iban (-) 

Samoa (— ) 

"Somali (— ) 

Karen (-) 

Subanum (— ) 

Tubatulabal (-) 

Lepcha (— ) 

Tupinamba (-) 

MinChia (-) 

Yoruba (-) 

Mossi (-) 

Riffians (-) 

Tallensi (-) 

Tanala (-) 

Thai (-) 

Thonga (-) 

Tiv (-) 

Animal husbandry societies. 


A second possible source of anxiety about being isolated and on 
one's own involves the subsistence economy. The relation of the sub- 
sistence economy to adult roles which demand independent and self- 
reliant behavior has been discussed by Barry, Child, and Bacon in 
a study of economy and child-rearing practices ( 1959) . They find 
that child-rearing practices stressing independence, self-reliance, 
and achievement are most typical of hunting and fishing societies, 
while child-rearing practices stressing obedience, responsibility, and 
nurturance are typical of societies with both agricultural and ani- 
mal husbandry. Societies with agriculture, and without animal hus- 
bandry, fall between these extremes. The correlation between the 
form of economy and a combined child-training measure of rela- 
tive "pressure for compliance" (composed of scores for obedience, 
responsibility, and nurturance training) versus "pressure for as- 
sertiveness" (composed of scores for independence, self-reliance, 
and achievement training) yields exceptionally strong coefficients 
of association of +.94 and +.93 for extreme and intermediate com- 
parisons (1959:59). This very high degree of association is thought 
to be due to the functional adjustment of child-rearing practices to 
the type of adult roles necessary to maintain food production. That 
is, societies with both agriculture and animal husbandry can best 
assure future food supply by "faithful adherence to routine" and 
therefore train children to be obedient and responsible, while in 
hunting and fishing societies individual initiative and skill is more 
adaptive, along with child-rearing practices stressing independence 
and self-reliance (1959:52). 

It is expected, then, that hunting and fishing societies will be 
likely to use dreams to seek and control supernatural powers, while 
societies with both agriculture and animal husbandry will be less 
likely to use dreams in this fashion. Societies with either agriculture 
or animal husbandry, but not both, should fall between these two 
extremes.^ This result is predicted for two reasons. First, according 
to Barry and co-workers, hunting and fishing societies place greater 
pressure on the adult to be independent and self-reliant. Second, 

■" Barry, Child, and Bacon group together both nomadic pastoral societies and societies with 
a combination of animal husbandry and agriculture, evidently considering the use of animals to 
be the crucial determinant in accumulation of food resources. However, the combination of 
agriculture with animal husbandry would be more likely to produce a stable and high food 
output than either economy separately. For this reason the groupings of categories of economy 
used by Barry and his co-workers have been altered slightly in this paper, and societies with 
animal husbandry and no agriculture have been put in the intermediate hunting and fishing group. 


hunting and fishing societies also place relatively greater pressure 
on the child to be independent and self-reliant. 

Table 2 presents the association between type of economy and 
the use of dreams to seek and control supernatural powers. The 
ratings on economy have been taken from Murdock ( 1957) . 

The results indicate that there is a strong and significant relation 
between the type of economy and the use of dreams. Approximately 
80 per cent of the hunting and fishing societies use dreams to seek 
and control supernatural powers, while only 20 per cent of the 
societies with both agriculture and animal husbandry use dreams 
this way. The intermediate societies, which have either agriculture 
or animal husbandry, but not both, fall between the two extremes, 
with 60 per cent of these societies using dreams to seek and control 
supernatural powers. 

Unfortunately, it is not possible to decide whether this association 
is due to the effect of child rearing, or to the effect of role pressures 
on adults. A separate test, using the child-training measure of pres- 
sure for compliance versus assertiveness, results in a significant cor- 
relation of assertiveness with an extensive use of dreams. However, 
attempting to control the effect of economy reduces this correlation 
drastically, although no firm conclusion can be drawn because of 
the large amount of overlap between type of economy and child- 
rearing practices. Attempts to use other measures of child rearing 
involving independence training, taken from Whiting and Child 
(1953), and unpublished scores rated by Barry and his associates, 
reveal a nonsignificant tendency for early indulgence of dependency 
and later severe socialization of dependency to go with extensive use 
of dreams to seek and control supernatural powers. 

Although economic conditions are related to the typical distance 
a son moves at marriage with the son moving further in hunting and 
fishing society, these two conditions seem to have clearly assessable 
independent effects. Within agricultural societies, the greater the 
distance between son and parent, the more likely a society is to use 
dreams to seek and control supernatural powers. The same relations 
hold within hunting, fishing, and pastoral societies. 

In general, the findings of this cross-cultural study support the 
notion that anxiety about being isolated and under pressure to be 
self-reliant may create an involvement with a type of fantasy about 
magical helpers. Both the use of fantasy and dreams, rather than 
ritual as the means of contact with the supernatural, and the use of 
personal helpers, rather than impersonal forces, seem to be involved 


in this complex. The type of economy and the degree of isolation of 
the married son from his parents have been found to affect this com- 
plex strongly, with hunting and fishing societies, and societies in 
which the son moves far away from his parents being more likely to 
use dreams to seek and control supernatural powers. Based on the 
rather weak correlations with child-training practices, and the lack 
of association with the isolation of the married daughter. I suspect 
that this effect is mediated by what happens to adults rather than 
children, and what happens to men rather than women.^ 

As a final summary, the following general conclusions about the 
relations between dreams, personality, and culture are tentatively 

1. There is a close association between dreams and the super- 
natural. This association consists of similarities between dream 
images and the conceptions of the supernatural, and also of the use 
of dreams to see and interact with the supernatural. This association 
does not necessarily indicate that dreams gave rise in the distant past 
to various conceptions of the supernatural, but would seem to indi- 
cate that similar psychological mechanisms may underlie both. 

2. There are a number of small bits of evidence to support the 
thesis that symbolism in dreams is a universal phenomena. If true, 
this means that man either innately or due to experience establishes 
a set of identities or equivalences without cultural tuition, and with- 
out awareness, and that these equivalences are in constant use. 

3. Dreams, it is assumed, can be used to reveal the dreamer's mo- 
tives. Further, the relation between the dream content and these 
motives may be more or less indirect and disguised. The most basic 
(and usually the most disguised) motives involve obtaining direct 
physical gratification from members of the nuclear family. These 
motives can be found in the dreams of people from all societies. 
The modal ways in which these motives are represented and de- 
fended against, however, vary culturally. 

4. It is also assumed that dreams have a cognitive as well as moti- 
vational component. The dreamer's waking life and, hence, his 
culture are represented in dreams. This representation is always dis- 

^ There is some evidence that early childhood conditions involving the identification process, 
whereby a young child comes to admire and wish to be like his or her parent of the same sex, 
also affects the use of dreams. It is thought that strong parental same-sex identification leads 
to fantasy about parent-like guardian spirits, and to the use of fantasy rather than ritual or 
acting out to relieve anxiety. This formulation is at present still tentative, and dependent 
upon further research. It may be that strong early same-sex parental identification is a neces- 
sary but not sufficient cause for a strong degree of cultural emphasis on dreams, with adult 
role stress involving isolation and independence a later "eliciting" factor. 


torted, however. Sometimes the distortion is mild, involving minor 
editing of material and bias in selection. At other times the dis- 
tortion may be drastic, involving complete reversal of normal ex- 
perience. Such distortion is probably due in part to the press of 
motivation, and especially conflicts in motivation. 

5. Culture may also specify the content which is appropriate to 
dreams under certain conditions. Where the individual is supposed 
to dream a certain dream, the retelling of these dreams is probably 
influenced by some degree of later elaboration. Acculturation may 
bring foreign material into such dreams, or completely break the 
pattern. The emotional reaction to dreams may be affected by the 
cultural definition of what is likely to take place in dreams, and in 
turn the cultural definition of the self may be affected by the kinds 
of events which occur in dreams. 

6. Dreams have numerous cultural uses. Prediction of the future 
and contact with supernaturals are the most common of these uses. 
Dreams are also used in native psychotherapies and as a means of 
selecting and rejecting personnel for various roles. One special use of 
dreams, to seek and control supernatural powers, seems to be caused 
by anxiety about being alone and needing to be able to be self- 
reliant. Societies in which the economy demands self-reliant be- 
havior on the part of the men, as in hunting, and societies in which 
the married son must move away from his natal family into another 
village are more likely to use dreams to seek and control supernatural 


Adelson, Joseph 

i960 Creativity and the dream. Merrill Palmer Quarterly 6:92-97. 

Barry, H., I. Child and M. Bacon 

1959 Relation of child training to subsistence economy. American Anthro- 
pologist 61:51—63. 

Benedict, Ruth 

1923 The concept of the guardian spirit in North America. Memoirs of the 
American Anthropological Association 29. 

Bourguigon, Erika E. 

1954 Dreams and dream interpretation in Haiti. American Anthropologist 

Dement, William 

1955 Dream recall and eye movement during sleep in schizophrenics and 
normals. Journal of Nervous and Mental Disease 122:263-269. 


Dement, William and N. Kleitman 

1957 The relation of eye movements during sleep to dream activity. Journal 
of Experimental Psychology 53:339—346. 

Devereux, George 

195 1 Reality and dream. New York, International Universities Press. 

1957 Dream learning and individual ritual differences in Mohave Shamanism. 
American Anthropologist 59:1036—1045. 

DiTTMANN, Allen and Harvey Moore 

1957 Disturbance in dreams as related to peyotism among the Navaho. Amer- 
ican Anthropologist 59:642—649. 

Eggan, Dorothy 

1949 The significance of dreams for anthropological research. American 
Anthropologist 51:171 — 198. 

1952 The manifest content of dreams: a challenge to social sciences. Amer- 
ican Anthropologist 54:469—485. 

1955 The personal use of myth in dreams. In Myth: a symposium, T. Sebeok, 
ed. Journal of American Folklore 68:445—453. 

196 1 Dream analysis. In Studying personality cross-culturally, Bert Kaplan, 
ed. Row, Peterson and Co., Evanston, III. 

Elwin, Verrxer 

1939 The Baiga. London, John Murray. 

Erikson, Erik H. 

1950 Childhood and society. New York, Norton. 

1954 The dream specimen in psychoanalysis. In Psychoanalytic psychiatry and 
psychology. Knight and Friedman eds., pp. 1 31-170. New York, In- 
ternational Universities Press. 

Firth, Raymond 

1934 The meaning of dreams in Tikopia. In Essays presented to C. G. Selig- 
man, Evans-Pritchard, Firth, Malinowski, and Schapera, eds., pp. 63-74. 
London, Kegan Paul, Trench, Trubner and Co. 

Freud, Sigmund 

1900 The interpretation of dreams, transl. New York, Basic Books, 1958. 
1920 A general introduction to psychoanalysis, transl. New York, Doubleday, 
Permabook edition, 1953. 

Goodenough, D., a. Shapiro, M. Holden, and L. Steinschriber 

1959 A comparison of "dreamers" and "nondreamers." Journal of Abnormal 
and Social Psychology 59:295-302. 

Griffith, R., O. Miyagi, and A. Tago 

1958 The university of typical dreams: Japanese versus American. American 
Anthropologist 60:1173-1179. 

Hall, Calvin S. 

1 95 I What people dream about. Scientific American, May. 

1953 The meaning of dreams, New York, Harper. 

1956 Current trends in research on dreams. In Progress in clinical psychology. 
New York, Grune and Stratton. 


Hallo WELL, A. Irving 

1938 Freudian symbolism in the dream of a Salteaux Indian. Man., 38:47—48, 
1955 The self and its behavioral environment. In Culture and experience. 
Philadelphia, University of Pennsylvania Press. 

HoBBEs, Thomas 

165 1 Leviathan. Reprint of ist Ed. Cambridge, England, University Press 
HoLMBERG, Allan R. 

1950 Nomads of the long bow: the Siriono of Eastern Bolivia. Smithsonian 
Institution: Institute of Social Anthropology Publication No. 10. 

HoNiGMANN, John J. 

1954 Culture and personaUty. New York, Harper. 

1 96 1 The interpretation of dreams in anthropological field work: a case 
study. In Studying personality cross-culturally, Bert Kaplan, ed. pp. 
579—585. Row, Peterson and Co., Evanston, 111. 
Hsu, Francis K. 

1959 Structure, function, content, and process. American Anthropologist 
Jacobs, Melville 

1959 The content and style of an oral literature; Clackamas Chinook myths 
and tales. Viking Fund Publications in Anthropology 26, New York, 
"Wenner-Gren Foundation for Anthropological Research, Inc. 
Jones, Ernest 

193 I On the nightmare. New York, Grove Press. Evergreen edition, 1959. 
King, Arden R. 

1943 The dream biography of a mountain Maidu. Character and Personality 
1 1:227—234. 
Kluckhohn, Clyde and "William Morgan 

195 1 Some notes on Navaho dreams. In Essays in honor of Geza Roheim, 
Wilbur and Muensterger eds. pp. 120-13 i. New York, International 
Universities Press. 

Lee, S. G. 

1958 Social influences in Zulu dreaming. Journal of Social Psychology, 47: 
Lincoln, Jackson S. 

1935 The dream in primitive cultures. London, Cresset Press. 
LowiE, Robert H. 

1924 Primitive religion. New York, Liveright. Black and Gold edition re- 
print, 1948. 
1922 The religion of the Crow Indians. Anthropological Papers of the Ameri- 
can Museum of Natural History 25:309-444. 
Mead, Margaret 

1932 An investigation of the thought of primitive children with special refer- 
ence to animism. Journal of the Royal Anthropological Institute 62: 
Murdoch, George P. 

1957 World ethnographic sample. American Anthropologist 59:664-687. 


Offenkrantz, W. and A. Rechtschaffen 

1960a Dream sequences of a patient in psychotherapy. Paper presented at the 
47th Annual Meeting of the American Psychoanalytic Association, At- 
lantic City, New Jersey, May. 
1960b Clinical studies of dreams in sequence. II Papers presented at meeting 
of American Psychoanalytic Association, New York, December. 
Radin, Paul 

1936 Ojibwa and Ottawa puberty dreams. In Essays presented to A. L. 
Kroeber. University of California Press. 


195 1 Experiments on symbolism in dreams. In The organization and pathol- 
ogy of thought, David Rapaport, ed. and translator, pp. 249-256, 
New York, Columbia University Press. 
RoHEiM, Geza 

1932 Psychoanalysis of primitive cultural types. International Journal of 
Psychoanalysis 13:2—224. 

1938 The nescience of the Aranda. British Journal of Medical Psychology 

1946 The Oedipus complex and infantile sexuality. Psychoanalytic Quarterly 

1947 Dream analysis and field work in anthropology. Psychoanalysis and the 
Social Sciences 1:87—130. 

1949 The technique of dream analysis and field work in anthropology. Psy- 
choanalytic Quarterly 18:471—479. 

1950 Psychoanalysis and anthropology. New York, International Universities 

ScHNEroER, David 

1 94 1 Aboriginal dreams. Masters thesis, Cornell University. 
Schneider, D. M. and R. L. Sharp 

(n.d.) Yir-yoront dreams. In manuscript. 
Sears, Walter E. 

1948 The Navaho and Yir-Yoront, their primitive dreams. Undergraduate 
honors thesis, Harvard University. 

Seligman, C. G. 

1924 Anthropology and psychology. Journal of the Royal Anthropological 

Institute 54:13—46. 
1927 Appendix. Religion and Art in Ashanti, Rattray, R. S. pp. 197—204. 

Oxford, Clarendon Press. 
1932 Anthropological perspective and psychological theory. Journal of the 
Royal Anthropological Institute 62:193-228. 
Stewart, Kilton 

195 1 Dream theory in Malaya. Complex 6:21-33. 
1954 Pygmies and dream giants. New York, Norton. 


1936 Dreams and dream interpretation of the Diegueno Indians of Southern 
California. Psychoanalytic Quarterly 2:195—225. 


Tresman, H., a. Rechtschaffen, W. Oefenkrantz, and E. A. Wolpert 
1 96 1? Studies in psychophysiology of dreams III. Relation among dreams in 
sequence. Archives of General Psychiatry (in press) . 
Wallace, Anthony 

1958 Dreams and the wishes of the soul: a type of psychoanalytic theory 
among the seventeenth century Iroquois. American Anthropologist 

Whiting, J. W. M. and I. L. Child 

1953 Child Training and personality; a cross-cultural study. New Haven, 
Yale University Press. 
Whiting, J. W. M. and R. G. D'Andrade 

1959 Sleeping arrangements and social structure: a cross-cultural study. 
Presented at American Anthropological Association Annual Meetings, 
Mexico City, December. 

chapter 1 1 



Northwestern University 

Rather than report upon a specific technique, this chapter will 
deal with some general methodological problems in relating theory 
to data. Rather than deal solely with an interdisciplinary specialty 
of "culture and personality," this chapter will emphasize the mu- 
tual relevance — at a methodological level — of anthropology and 
psychology for each other. This relevance is believed to hold even 
when each discipline is focused upon its own pure problems, as well 
as when they enter into interdisciplinary collaboration. This mutual 
methodological relevance is emphasized as a mode of contact sepa- 
rate from the inevitable mutual relevance of their substantive 
theories. The latter, while more important, has also received more 
repeated attention, and is in any event not the topic treated here. 

Anthropology os a Source of Discipline for Psychological Theory 

There is no need to reiterate or to document here the tremendous 
influence which anthropology's culture-personality studies have 
had upon social psychology since the 1930's. From the tenor of 
some of the papers of this volume and from other professional stock- 
takings by anthropologists (for example, Bennett 1946, Kluckhohn 
1954b, Honigmann 1954), it can be gathered that many anthro- 
pologists feel somewhat uneasy about this very great popularity of 
what may be a not-too-dependable product; that many might ex- 
plain the rapid diffusion of this trait complex more as due to the 
extreme needs of the new converts than to the efficacy of the in- 



vention, that is, an acceptance phenomena more akin to the diffu- 
sion of the Ghost Dance Rehgion than to the spread of the com- 
pound bow, barbed fishhook or better mousetrap. As an academic, 
experimentally oriented, and methodologically anxious social psy- 
chologist, I, of course, share these misgivings. However, even when 
in an incomplete and fragmentary form, anthropological evidence 
has served as a source of discipline, as well as a source of inspiration 
to psychological theory. 

The first, and perhaps still most needed influence is at a very gen- 
eral level. This is the message of cultural relativism. While recog- 
nizing that anthropologists themselves are not too happy with this 
slogan, and that the perspective may not be adequate for anthro- 
pology's theoretical purposes, the message it has to offer is still very 
much needed by academic psychologists. Implicitly, the laboratory 
psychologist still assumes that his college sophomores provide an 
adequate base for a general psychology of man. (Such assumptions 
of universality are automatic for any provincially enculturated 
ethnocentric.) For social psychology these tendencies have been 
very substantially curbed through confrontation with the anthro- 
pological literature. Continued confrontation, however, will be 
required to prevent relapse. For the general psychologist, most of the 
message is yet to be learned. 

The message of cultural relativism is very general and nonspecific. 
Often it is merely a general caution against intemperate generali- 
zation. (And often it takes the extreme of a negativistic denial of 
the possibility of any generalization.) The central purpose of this 
paper is to call attention to more concrete and specific methodologi- 
cal relevance. As Flonigmann (1952, 1954), Whiting (1954), and 
Child (1954) have pointed out, anthropological evidence has been, 
and can continue to be, of invaluable service as a crucible in which 
to put to more rigorous test psychology's tentative theories, enabling 
one to edit them and select among alternatives in ways which labora- 
tory experiments and correlational studies within our own culture 
might never make possible. 

While this can never be anthropology's central role, what is here 
argued is that anthropology provides an important part of the scien- 
tific apparatus of psychology, particularly for personality theory. 
This is said within a perspective upon the strategy of science which 
sees experimentation and the other methods of science as having es- 
sentially an editorial function. That is, scientific data serve to choose 
among, prune out, and in this sense, edit theories. Essential to build- 


ing a science are such laboratories. Where all are lacking, no science 
is possible. In the absence of the possibility of experimentation with 
modes of child rearing and personality formation, a science of per- 
sonality would be all but impossible were it not for the "laboratory" 
of cross-cultural comparison opened up by the anthropologist. 

To illustrate this role, several condensed and oversimplified exam- 
ples are offered. Note that these are organized around problems in 
psychological theory. (That such problems are not central to an- 
thropology should not distract us from this important service.) 
Though the "facts" in the illustrations may in fact be controversial, 
it is hoped that they exemplify the possibility, if not the actuality, 
of the editing role of anthropological data. 

1. Freud validly observed that boys in late Hapsburgian Vienna had hostile 
feelings toward their fathers. Two possible explanations offered themselves — the 
hostility could be due to the father's role as the disciplinarian, or to the father's 
role as the mother's lover. For reasons that can be neglected here (but see Bakan 
1958) Freud chose to emphasize the role of the mother's lover. However, work- 
ing only with his patient population there was no adequate basis for making the 
choice. The two rival explanations were experimentally confounded, for among 
the parents of Freud's patients the disciplinarian of little boys was usually the 
mother's lover, (Remember that in Freud's day it was the morality of one's par- 
ents more often than their immorality that drove one to choose the analyst's 
couch over other couches, so that Freud got a biased sample.) Malinowski (1927) 
studied a society in which these two paternal roles were experimentally disen- 
tangled, in which the disciplinarian of young boys and the mother's lover were 
not one-and-the-same person. And in this society, the boys' hostility was addressed 
to the disciplinarian, not to the mother's lover. This outcome makes the Oedipal 
hostility more easily encompassed within the framework of a simple hedonistic 
learning theory such as that of Thorndike or Hull. While the love-jealousy and 
the punishment Oedipal theories are no doubt both appropriate to some extent, 
Malinowski's work helps to integrate personality theory within learning theory 
and gives us a firmer base upon which to predict the Oedipal complex of the 
son of a commuting suburban father where the mother is the only source of 

2. Pettitt's (1946) monograph on educational practices among North Ameri- 
can Indian tribes serves the purpose of calling attention to the fact that our 
theories of learning and cognition predict trouble for the modern emancipated 
American family. According to learning experiments, conditioned fear and con- 
ditioned hostility are the unrational product of temporal contiguity between 
stimulus and pain, or between stimulus and frustration. And if we go to cognitive 
psychology, we find that the perception of causality, and with this the phenomenon 
of blaming, are likewise functions of temporal and spatial contiguity (Heider 
1944, Michotte 1946). From these theories it follows that in a society such as 
intellectual suburbia, where the parents stand alone in representing the restraints 
which society passes on to children, the parents will become the stimuli for condi- 
tioned hostility on the part of the children, the children will perceive the parents 


as causing, as to blame for, their frustrations. Thus, the conditioning and/or the 
causal perception processes predict a chronic divisive force within the modern 

With the inevitable selective process in which, among the countless customs 
that are tried, some are preserved more readily than others (e.g., Keller 193 i), 
one can expect that in stable societies preventive customs will have grown up 
around this inevitable parental-resentment problem. Pettitt's (1946) analysis 
spells out the role of shamans and kachina dancers as disciplinarians, of the 
avunculate, of age grade systems, all as devices serving to deflect the discipline- 
induced hostility of the child away from the parent, and, thus as preserving 
intrafamilial solidarity. Reading his monograph gives one both a greater apprecia- 
tion of the relevance of learning theory for predicting intrafamilial attitudes, and 
parenthetically a greater sympathy for those unsophisticated parents in our own 
culture who attempt a similar deflection of childish hostility away from them- 
selves through invoking the sanctions of the policeman, the boogeyman, Santa 
Claus, or a reified God. (On the other hand, perhaps it is well that in our cul- 
ture the socialization-induced hostilities are associated with parents, for our occu- 
pational structure requires new entrants to the labor force who are willing and 
eager to leave home permanently. Just such a labor force is lacking in some of 
the underdeveloped countries, perhaps in part because of the greater "wisdom" 
of their intrafamilial relationships.) 

3. Every practicing psychoanalyst doing therapy with parents has probably 
recognized that the parent contributes much of the irrational and projected 
attitudes that comprise the intergenerational Oedipal interaction — yet this recog- 
nition is little represented in the literature, although not totally absent (e.g., Hsu 
1940, Wellisch 1954). Recently the Herskovitses (1958a, b) have not only called 
attention to the ubiquity of the theme of the father's hostility toward his first 
born son in the Oedipus-type myths of Africa and Eurasia, but have in addition 
hypothesized that this paternal hostility to the newborn represents a reactivation 
of the father's sibling-rivalry hostility, acquired in his childhood in reaction to 
a younger sibling who abruptly displaced him in the total attention of the mother. 
The Herskovitses came to this hypothesis working with the mythology of Da- 
homey, a polygynous society in which each wife has her own hut, and in which 
a newborn child is continually with the mother, at work during the day and on 
the sleeping mat at night, until at around the age of two or three it is displaced 
by a younger sibling. Corresponding to this familial pattern is a mythology 
exceptionally full of strife between brothers and between generations, and in 
which the older brother or the older generation is portrayed as the initiator of 
the hostihty. 

Once pointed out, this seems exactly what one would expect from considera- 
tions of stimulus equivalence and habit transfer. Certainly in many cultures be- 
sides Dahomey (e.g., Levy 1935, Paul 1950, Henry 1944, Spiro 1953) hostility 
toward younger siblings is among the most characteristic and strongest learnings 
of childhood. When later as an adult an older sibling is presented with the new 
stimulus that his own child constitutes, this novel stimulus can be expected to 
elicit the strongest of the response tendencies learned in the past toward a similar 
stimulus, that is, the responses learned toward the younger sibling as an infant. 
The degree of this projected hostility would presumably be correlated with the 


degree to which the child in its first years had the undivided attention of the 
mother, and, hence, was the more frustratingly displaced at the end of the infancy 
period. If initiation rites be taken as symptomatic of the hostility of the older 
generation toward the younger (as Wellisch 1954 has plausibly interpreted in- 
fanticide and the sacrificing of children to be), then one might expect the high 
correlation between length and degree of infant monopoly of the mother's atten- 
tion and the hostility of initiation rites, which Whiting and co-workers (1958) 
report. (Initiation rites could also, on the basis of the same theory, represent the 
still more direct expression of the hostility of the older already initiated brothers 
toward the younger.) It can be noted that Dahomey is included in Whiting's sam- 
ple, and is scored in the very highest category for severity of initiation rites. In 
Kwoma, the child's displacement may be through the father's return to the sleeping 
mat, but this is not the pattern in Dahomey. In general, displacement by a younger 
sibling is probably the more usual mechanism. If the projected sibling hostility 
is a relevant part of the explanation, then upon examination we should find both 
actual and mythological sibling strife more prevalent both in cultures with the 
harsher initiation rites and in the cultures with the longer infant monopoly of 
the mother's attention in infancy. 

I have recently confirmed the stimulus equivalence of offspring and younger 
siblings assumed in this derivation in an unpublished study of the types of con- 
fusions of names that occur on the part of parents of college sophomores: When a 
parent mistakenly calls a child by the name of one of the parent's own siblings, 
the name of a younger sibling (of the same sex as the child) is most frequently 
involved. This study can not, of course, confirm the hostility aspects of the inter- 
pretation. Note also that this theory predicts a relative absence of parent- 
originated hostility for parents who were only children or youngest in their 
families, except insofar as the newborn is a genuine displacer of the parent in the 
attentions of the spouse. 

4. Freud presented psychology with an insightful, but doubly double-jointed 
theory relating drive fixation in childhood and adult behavior. On the one hand, 
the fixation could be produced by overindulgence of the drive in childhood, or 
by its opposite, underindulgence. As to expression in adult life, fixation could 
express itself in excessive preoccupation with drive-relevant things or by its 
opposite, a counterphobic avoidance. Such a prediction is somewhat more specific 
than no prediction at all, but when combined with the inevitable errors of clas- 
sification, the polar-cross scatter diagram which it predicts may not be distin- 
guishable from a zero correlation. And whereas on many points, psychoanalysis 
and hedonistic-associationistic learning theories agree, the learning theories predict 
most easily a parallelism between conditions of acquisition and those of expression 
and transfer, rather than compensatory or complementary relationships, since 
memory but not energy storage is expected to persist. Whiting and Child's ( 1953 ) 
study may be interpreted as confirming those aspects of the Freudian hypothesis 
which thus agree with the learning theory interpretation. Persons for whom a 
given drive had been associated with frustration in childhood show phobic re- 
actions regarding it in adult life (negative fixation). And insofar as infantile 
indulgence and gratification had adult symptoms, those who found a given drive 
a source of gratification in childhood sought it out as a source of cure in adult 
life. Here, again, the result has been in the direction of integrating personality 


theory with learning theory. Here, again, the anthropological data have been 
efficacious in selecting among alternative psychological hypotheses. And as Child 
(1954) shows, insofar as relationships, Freudian or otherwise, have been estab- 
lished between early child training and adult behavior, the confirmations have 
come primarily from the studies of cross-cultural breadth, rather than from 
studies making use of the small range of differences within our own culture. 

5. Other studies using the cross-cultural method seem to confirm the positive 
transfer of attitudes between childhood reinforcement conditions and adult per- 
sonality, the assumptions of stimulus equivalence, transfer, displacement in 
approach-avoidance conflicts, and so forth. Spiro's (1953, 1958) demonstration 
of the parallel between infant training by parents and attitudes toward spirits 
is interpreted as confirmatory in this regard. This may seem contradictory, since 
in his 1953 paper, Spiro takes his evidence as justifying a choice in favor of a 
perceptual rather than a learning theory. As I understand it, learning theories 
are silent as to the nature of conscious contents. Hence, evidence regarding con- 
scious contents are not contradictory to learning theory. In particular, evidence 
regarding "perceptions of" objects cannot be interpreted as corresponding to the 
stimulus terms of learning theory. Usually a better translation of "perceived as" 
is "responded to as to." On this ground, learning theory expects the authority 
symbols of adult life to be responded to as (to be perceived as) were the au- 
thority figures of childhood to which the responses (perceptions) were originally 
learned. For more details on this mode of integrating theoretical terminologies, 
see Campbell ( 1961 ) . For more evidence on the parallels between attitudes toward 
parents and toward spiritual beings, see Lambert and co-workers (1959). 

In general, the evidence of social anthropology is seen as having 
a salutary and disciplining effect upon personality psychology, serv- 
ing, paradoxically, to make personality theory more clearly a part 
of the learning theory of general psychology. 

Some Psychological Comments on Anthropological Method 

It is probably true that the testing of psychological theories must 
remain a very minor part of the research agenda of the anthropolo- 
gist. In addition, the great difference in task must be recognized be- 
tween the descriptive, humanistic task of one who seeks to record 
all aspects of a specific cultural instance and the task of the abstrac- 
tive and generalizing "scientist" who wants to test the concomitant 
variation of two isolated factors across instances in general. Co- 
operation between these orientations is often difficult — but is helped 
rather than hindered by the explicit recognition of the great differ- 
ence in goals: Too often those in one camp regard those in the other 
as the willful practitioners of a wrongheaded approach, implicitly 
assuming a common goal. Both orientations are represented in the 
present volume, in some instances both within a single person. The 
descriptive-humanistic rather than abstractive approach has in the 


past been typical of much of anthropology. On the other hand, 
Honigmann (1952), Whiting (1954), and Spiro (1953, 1958) 
have presented the abstractive, hypothesis-testing commitment. 
Murray (1949) and Gillin ( 1954) have called for such an orienta- 
tion in previous symposia on culture and personality. My interests 
are wholly of this sort, and some of the methodological comments to 
follow are thus irrelevant to the more typically descriptive anthro- 
pological undertaking. Many of these comments come from an in- 
terest in a potential psychology of induction (Campbell 1958a, 
1959) , and in particular from an application of knowledge about 
human perception, learning, and biases to the calibration of the hu- 
man observer as a scientific measuring instrument (for example, 
Campbell, Hunt, and Lewis 1957, 1958, Campbell 1958b) . 

Before going into these details, it may be well to note a common 
cause joining the abstractive-generalizing orientation central to this 
paper and the descriptive-humanistic orientation as it has been 
modally represented in anthropological research training. Both 
stand in opposition to the undisciplined generalizations often found 
in the more dramatic efforts to interpret man and culture. Both 
look askance at the sweeping generalizations of a Spencer, a Speng- 
ler, a Toynbee, or a Nietzsche when offered as established scientific 
truth. This common ground is not always noted, and, indeed, each 
orientation tends to attribute undisciplined generalization to the 

In the major departments of anthropology of the 1920's and 
1930's the theoretical excesses of a previous generation of anthro- 
pologists led to an emphasis upon objectivity in field work which 
was antitheoretical insofar as adherence to theory had in the past 
served to reduce the objectivity of field work. Herskovits (i960) 
has recently called attention to a superior objectivity for the hu- 
manistic aspects of anthropological study. Both the descriptive- 
humanistic orientation and the abstractive, hypothesis-testing 
orientation wish to avoid self-deception and bias in the data collec- 
tion process. Both call for reliable, intersubjectively communicable 
observations. Both are ideally hardheaded, skeptical, modest, and 
conservative in their orientation to factual knowledge. For these 
reasons, many of the topics covered in what follows are of joint 
relevance. This point is made without weakening the appeal for the 
mutual recognition and respect for a separateness of task and divi- 
sion of labor between the two orientations, both of which are essen- 
tial in the complete study of man. 


The Relation of Intersubjective-Verifiability to Directness 
of Sense Receptor Access 

It goes without saying that a science of either type cannot be built 
without intersubjective verifiabihty of observations. Psychological 
research on the accuracy and person-to-person agreement in inde- 
pendent reporting seems summarizable by the statement that the 
greater the direct accessibility of the stimuli to the sense receptors, 
the greater the intersubjective verifiabihty of the observation. The 
weaker or the more intangible, indirect, or abstract the stimulus 
attribute, the more the observations are subject to distortion. 

It is quite conceivable that there are some aspects of culture, in- 
cluding its over-all pattern or ethos, that are so abstract or indirectly 
inferred that intersubjective verifiabihty is lost. If this is so, then 
until corrected, these aspects cannot become a part of science, and 
we, as scientists, should concentrate on those aspects upon which we 
can get agreement. Recently Holmes (1957, 1958) has reported a 
restudy of some of Mead's work on Samoan society, which along 
with the other restudies of recent years (such as Li An-Che 1937, 
Bennett 1946, and Lewis 195 1) supports the methodological ex- 
pectation of greater verifiabihty to the more palpable and visible. 
As far as the great bulk of Mead's ethnology. Holmes confirms her 
findings, stating "the reliability of Mead's account is remarkably 
high." While he reports some differences in the description of tradi- 
tional political systems and other matters, on matters of material 
culture and observable custom, there is general agreement. This ex- 
tends also to the observed absence of an adolescent disturbance on 
the part of the girls, and the easy transition from childhood to adult 
life. But upon several of the broader aspects of ethos, his findings 
are in complete disagreement, for example, upon the lack of special- 
ized feeling in human relations, the lack of competitive spirit, the 
lack of crisis in human relations, and the importance of "Mafau- 
fau," or the gift of wise judgment. In the context of his presenta- 
tion, one cannot easily interpret these differences as due to culture 
change in the intervening years, but rather one must interpret them 
as disagreement in the description of aspects of "the same" culture. 
If, as Mead has said, "in the matter of ethos, the surest and most 
perfect instrument of understanding is our own emotional response" 
( Mead and McGregor 1951:300), then ethos may indeed be beyond 
the realm of scientific study. This lack of intersubjective verifiabil- 
ity is not inevitable however. In the methodological pattern of 
Whiting and Child (1953) some of the relevant data are made 


much more directly accessible to the senses; in addition, the integra- 
tive patterning is made a matter of explicit public combinational 
formulas. Through the use of methodological procedures developed 
to control the demonstrated biases of human observers, judgments 
of the intangibles of ethos may be made intersubjectively confirm- 
able, demonstrable in reliability studies. 

Adaptation Level and Contrast Effects 

In considering the faults of our laboratory experiments in social 
psychology, we have come up with a list of recurrent flaws, some of 
which also apply to other types of data collection. One of these has 
been called infelicitously "instrument decay" (Campbell 1957) : 
When human observers are used as the measuring device their judg- 
mental standards often change in ways that may be misinterpreted 
as experimental effects. A major source of such "instrument decay" 
is a set of phenomena in human judgment summarized by Helson 
(1947) under the concept of "level of adaptation." Its role in social 
science field work may be illustrated by the anecdote in the follow- 
ing paragraph.^ 

In the last several years, considerable numbers of Russian experts 
from American universities have been sent on visits to the U.S.S.R. 
In part, they have had different itineraries, some going first to Len- 
ingrad, others first to Moscow, and so forth. In comparing notes 
later they have found themselves in disagreement as to which Rus- 
sian city (Leningrad or Moscow) was the more drab and which the 
more lively. These differences in opinion have turned out to be 
correlated with the differences in itinerary: which ever city one 
visited first seemed the more drab. Against the adaptation level 
based upon experience with familiar United States cities, the first 
Russian city seemed drab and cold indeed. But a stay in Russia modi- 
fied the adaptation level, changed the implicit standard of reference 
so that the second city was judged against a more lenient standard. 
Such a process is what would be predicted by extrapolation from 
laboratory and field studies of the effect of context upon clinical 
psychology judgments (e.g., Campbell, Hunt, and Lewis 1957, 
1958). Of course, other processes were also involved — familiarity 
with the Russian vernacular, sensitivity to the expressive compo- 
nents of voice tone and gesture, and other skills facilitating warm 
social contacts were increasing. All such effects were operating, 
however, to change the calibration of the human observer, and thus 
to bias his reports in a systematic way. 

^ I am indebted to Professors Derning Brown and Raymond Mack for this information. 


How can we learn of, and correct, such bias? The anecdote is in- 
structive in this regard. This bias would not have been noted if all of 
the visitors had had the same itinerary. Their actual pattern consti- 
tuted a counterbalanced, observational schedule, and could have 
been analyzed as a crossover design (Cochran and Cox 1950) to 
determine the main effects of firstness versus secondness, of city, 
and of observer. Essential in the control were multiple observers and 
multiple sequences. 

Today many anthropologists, as in Africa, are combining basic 
ethnography with acculturation studies, and are faced with the de- 
cision as to whether to study first the members of the tribe who re- 
main in the bush, or the members living in the westernized city. 
Combining the principles of adaptation level with other principles 
of bias, particularly those involving assimilation errors or transfer 
(see Campbell 1958b for a survey of such biases) some predictions 
can perhaps be made: (i) If one compares anthropologist's impres- 
sion of the indigenous bush culture under the two orders (bush-city 
versus city-bush) , this indigenous culture would probably appear 
more strange and exotic under the bush-city order. This is because, 
under that order, the bush culture is perceived with a more diver- 
gent adaptation level than that provided when the partially west- 
ernized members of the culture have been previously studied in the 
city. (2) The bush data might be better in detail and intimacy of 
records for the city-bush order than for the bush-city order. This 
might be expected insofar as rapport is increased by the familiarity 
with the culture and the friendship bonds acquired through the 
city fieldwork with the partially acculturated members of the eth- 
nic group. (3) The observation of "survivals" of the indigenous 
culture among the westernized urban descendents is no doubt en- 
hanced by detailed knowledge of the relatively untouched bush cul- 
ture. Thus, such "survivals" might be noted in greater number in 
the bush-city order. These predictions cannot, of course, be made 
unequivocally. But whatever the direction predicted, there are ade- 
quate grounds to expect the two sequences to produce different 
results, particularly on those intangible matters most relevant to the 
culture-personality problem. 

The source of error is great enough, and a considerable remedy is 
near enough at hand, so that we are morally bound to request from 
our sources of financial support the funds to implement them — par- 
ticularly since all concerned should now recognize how precious to 
the social sciences is our rapidly dwindling supply of novel and 


independent social systems. The cheapest remedy would be to 
schedule the field work so that it was broken up into several alter- 
nating visits to each location, bush and urban, allowing boih con- 
ditions to be recompared several times, and both to be judged against 
the end-of-field-trip adaptation level. This could probably be ac- 
complished with lo per cent increases in the travel budgets and 50 
per cent increases in the field residence budgets — certainly not im- 
possible to promote once the importance is recognized. A more 
complete control would double field costs by having the field- 
workers work in pairs, one starting in the bush and one in the city, 
and trading locations from time to time. This approach would also 
offer an important control over the "personal equations" or idio- 
syncratic predilections of the observers, biases of a more permanent 
and less predictable sort than those due to adaptation level. There 
would seem no doubt but that this additional cost would be justified. 

Adaptation Level and Usable Vocabulary 
in Cross-Cultural lntervie\/ing 

One of the emphases of the present paper is upon the desirability 
of some studies which collect data on a limited set of topics from 
many cultural units. This is advocated not as a substitute for the 
intensive ethnography of single peoples, but rather as a needed addi- 
tional mode of data collection, particularly for those correlational 
types of analysis in which dozens of cultures are needed. In such 
multiple-culture studies the field work would be particularly de- 
pendent upon interviews with informants, the anthropologist him- 
self not having time to observe directly all of the customs about 
which he inquired. In such studies the phenomenon of adaptation 
level creates for a class of descriptive words "translation" problems 
over and above the troublesome fact of language differences. That 
is to say, these adaptation-level problems would remain even if the 
heterogeneous cultures were to "speak the same language" as the 

The words or concepts in question are those used to characterize 
the tribe as a whole which imply degrees of departure from a usual 
norm or adaptation level, this norm being itself provided by the 
average behavior or experience of the tribe itself. Such words are 
usable to denote individual differences within the tribe, but not to 
characterize over-all attributes of the culture. Thus, for a hypo- 
thetical "wholly isolated" tribe, lacking a range of other peoples 
for comparison, one could not interpret for cross-cultural compari- 


sons answers to questions such as: "Are your people happy, intelH- 
gent, hard working, strict with children, warm, friendly, prudish, 
joking, able to endure pain, and so forth?" 

Anthropologists, experienced with many cultures and having a 
common base in European cultures, may be able to make such ob- 
servations and judgments reliably, particularly if the fluctuations 
of their own adaptation levels, as described above, be compensated 
for. But a completely isolated tribe would have no "lingua franca," 
no intertribal measuring stick against which to calibrate their use 
of the terms. And even though informants might reliably employ 
the frame of reference provided by the several adjacent tribes, given 
the ubiquitous tendencies toward regional similarity, this would not 
entirely eliminate the problem. 

For some of these topics, modes of questioning are available which 
may avoid this problem. Such questioning may make use of in- 
ternal comparisons within the tribe ("Are children happier than 
adults?") . More typically, the problem may be solved by reducing 
the question to sample behaviors from the implied syndrome, em- 
ploying terms referring to qualitatively discrete and universal be- 
haviors: "Upon what occasions do women smile and laugh." "What 
does a mother do when her child cries?" "What are the times during 
the day when a man works — or rests?" These suggestions, however, 
do more to raise the problem than to suggest a solution. 

The Uninterprefabllity of Comparisons between 
But Two Noturol Instances 

In view of the importance of Malinowski's challenge to the love- 
jealousy interpretation of the Oedipal conflict, it is unforgivable 
that his observations have not been replicated. However thorough 
his field work on other points, his published evidence on this point 
is very thin indeed. While he alludes to evidence from manifest 
dream content, of the type that Dorothy Eggan (1952) has dis- 
cussed, what we need are substantial samples of detailed records of 
the dreams of boys and girls and men and women. 

But while there is a crying need for verifying and extending 
Malinowski's evidence on Trobriand intraf amilial attitudes, such a 
replication is of minor importance for testing the Freudian hy- 
pothesis. We who are interested in using such data for delineating 
process rather than exhaustively describing single instances must 
accept this rule: No comparison of a single pair of natural objects 
is interpretable. Between Trobriand and Vienna there are many 


dimensions of differences which could constitute potential rival 
explanations and which we have no means of ruling out. For com- 
parisons of this pair, the ceteris paribus requirement becomes un- 
tenable. But data collection need not stop here. Both the avunculate 
and the European arrangement are so widely distributed over the 
world that if testing Oedipal theories were our purpose, we could 
select a dozen matched pairs of tribes from widely varying culture 
areas, each pair differing with regard to which male educates and 
disciplines the boy, but as similar as possible in other respects. As- 
suming that collections of dreams from boys showed the expected 
differences between each pair, then the more such pairs we had, the 
fewer tenable rival hypotheses would be available and, thus, the 
more certain would be our confirmation. 

There is an analogous ceteris paribus problem with the use of a 
single measuring instrument. An established difference between two 
matched populations on a single questionnaire item is likewise un- 
interpretable because there are so many rival hypotheses to explain 
the difference — the groups may differ because of their reactions to 
the first word, or to the second word, or to the grammatical features 
of the wording rather than the semantic features, and so forth. 
However, if there are multiple indicators which vary in their irrele- 
vant attributes, and if these all agree as to the direction of the dif- 
ference on the theoretically intended aspects, then the number of 
tenable rival explanations becomes greatly reduced and the confir- 
mation of theory more nearly certain (Campbell 1957:3 lo, Camp- 
bell 1959, Campbell and Fiske 1959). Doob (1958) has recently 
demonstrated the seriousness of this problem in cross-cultural stud- 
ies, in an important paper which should be read by every graduate 
student planning to do research on culture and personality. On this 
point, it has been psychologists studying college sophomores and not 
anthropologists who have been most guilty of a naive overdepend- 
ence upon single instruments, and our critical literature on "re- 
sponse sets" (e.g.,Cronbach 1946, 1950, Chapman and Bock 1958) 
shows how misleading this can be. 

The Whiting and Child Studies 

From this sample of content interests and methodological biases, 
it will come as no surprise to learn that I regard studies of the Whit- 
ing and Child type (Horton 1943; B. B. Whiting 1950; Murdock 
and Whiting 195 1; McClelland and Friedman 1952; Whiting and 
Child 1953; Whiting 1954; Wright 1954; Barry 1957; Barry, 


Bacon and Child 1957; Freeman and Winch 1957; Rose and Wil- 
loughby 1958; Whiting, Kluckhohn and Albert 1958; Spiro and 
D'Andrade 1958; Lambert, Triandis and Wolf 1959; Barry, Child 
and Bacon 1959; Whiting 1959) as very important steps toward a 
science of personality and culture, as well as one of the major events 
in the social sciences of the past twenty years. It can be seen why, 
from this perspective, the earlier studies of individual cultures such 
as Trobriand, Dobu, Kwakiutl, Kwoma, If aluk, and Brobdingnag, 
can be regarded more as sources of hypotheses than as confirming 
evidence for the purposes of a science of culture and personality. 

This is stated so strongly because it is felt that until very recently 
anthropology has in general both rejected and neglected these stud- 
ies, and that the reasons for this rejection might well be discussed. 
These reasons have been given little attention in the anthropological 
journals. The neglect has been so great that it is difficult to docu- 
ment the rejection. The few published references in anthropological 
publications are in general favorable (Gladwin 1954, Kluckhohn 
1954, Honigmann 1952, Spiro 1958) . The neglect is perhaps indi- 
cated by the fact that until Spiro 's (1958) study appeared, none of 
the dozen or so prior studies had been presented in an anthropologi- 
cal journal. Thus, for the details of the rejection, the writer will 
have to depend for the most part upon informal sampling of the 
opinions of anthropology graduate students and faculty members 
at some seven universities, relying primarily on their reports as to 
how "anthropologists in general" felt. These are the objections heard 
most frequently (Gladwin 1954 and Spiro 1958 mention several 
of them): 

1. This is not anthropology. This objection can be, of course, an entirely legiti- 
mate expression of differences in goals. It may reflect upon the fact that problems 
of psychological theory rather than anthropological theory are under test. It can 
express a commitment to anthropology's task, comparable to that of the historian, 
of documenting in detail the full complexity of single instances. But this objec- 
tion is usually a concomitant of other objections which reject the studies for the 
abstracting-generalizing purpose also. 

2. Taking fragments of a culttire and attempting to interpret them apart from 
the tvhole cultural complex is impossible or illegitimate. Spiro (1958) has cited 
this widespread objection, and has correctly called it an empirical question to be 
answered by the final outcomes of trying the approach. Such criticisms may be 
right. It may be that none of the findings will stand up under cross-validation, 
that no correlational laws relating aspects of cultural phenomena can be estab- 
lished. Such laws cannot be ruled out on a priori grounds, however. 

From the standpoint of an empirical science of induction (Campbell 1959), 
it must be expected that there may be many problem areas in which a science 


cannot be established. In the terminology of the analysis-of-variance statistics 
of experimentation, if in a given area one always finds significant highest-order 
interactions, and never finds significant main effects or lower order interactions, 
then a science probably never can be developed. The healthy infancy of the suc- 
cessful sciences seems to have been predicated upon the stimulating nourishment 
of crude but effective ceteris paribus laws. For example, the force fields of atomic 
nuclei extend in infinite distance in all directions. However, they decay so rapidly 
as a function of distance that they can be disregarded in the statement of many 
crude laws, such as those embodied in Archimedes' mechanics. Were this not so, 
were Archimedes to have had to Umit himself to statements about each particular 
instance, then physics never could have developed. The critics of the generalizing 
social scientists are right in cautioning against claiming effective ceteris paribus 
laws when one hasn't got them, but pointing to the obvious idiosyncracy of every 
person, tribe, or swinging cathedral chandelier provides no a priori basis for re- 
jecting the enterprise. 

3. The data in the Hziman Relations Area Files and in the research monographs 
available are inadequate to the purpose. While it is obvious to every one, "Whiting 
and Child first of all, that better data would be desirable, the incompleteness and 
the inaccuracy of the files cannot explain away the striking correlations obtained. 
Error of this sort loivers correlations, rather than raises them. Significant high 
correlations can be explained away as due to the incompetence of the ethnography 
only if a systematic source of error be found to be confounded with the classifica- 
tions used — if, for example, all of the indulgent cultures turned out to have been 
described by French anthropologists and all of the high socialization anxiety cul- 
tures by German ethnologists. Such systematic sources of error have not been 
suggested and are extremely unlikely. 

4. A specific tribe has been misclassified, or there is another tribe which they 
don't report upon which doesn't fit. The abstractive-generalizing social scientist 
knows that in dealing with natural groups ceteris are not in fact paribus, and he 
therefore expects exceptions which represent the operation of many other laws 
which he as yet knows nothing of. Such exceptions are repeatedly found in the law- 
confirming scatter diagrams of biology and psychology. If the over-all significant 
relationship still persists when the specific errors are corrected and the new cases 
plotted, the exceptions are not invalidating. 

5. The process of coding qtialitative data into numerical categories offers op- 
portunities for a subjective bias tchich generates the correlations. This criticism 
is certainly occasionally valid, and may explain away the results of one striking 
relationship (McClelland and Friedman 1952, as restudied by Child et al. 1958). 
The basic Whiting and Child studies have been, however, scrupulously careful 
about this. They may have more trouble on this score in their new studies with 
their specially trained fieldworkers who can hardly remain in ignorance of the 
hypotheses under test. 

6. Many correlation possibilities have been inspected and only those that are 
high reported; thus, high valties may be due to chance even if apparently sta- 
tistically significant. This criticism is in some degree appropriate to most ex- 
ploratory studies that admit of reformulating hypotheses in the course of the 
investigation. It can be answered only by testing the relationships on new sam- 
ples, and this, of course, should be done. (The social sciences differ from the 


physical sciences in lacking the voluminous replication research that validates 
and revalidates every important new discovery. ) 

7. Since cultures are not independent, the usual tests of significance are not 
appropriate. I am not competent to enter into the abstruse statistical considera- 
tions that are involved here, but do want to point out some more common sense 
considerations. The criticism applies equally to samplings of persons and their 
response dispositions, where we normally use tests of significance without qualms. 
Cluster-sampling techniques (e.g., Kish 1956) are appropriate for computing a 
more accurate and larger error term. The criticism would be particularly damning 
if it turned out that regional areas were confounded with theoretical classifica- 
tions: if, for example, most of the indulgent cultures came from the South Seas 
and most of the high socialization-anxiety cultures from Africa. This has not 
been the case, however. Furthermore, when Whiting and Child (1953:168) 
analyze their data so as to show that a given relationship holds within each of five 
major culture areas, the use of a number of tribes from each of several culture 
areas becomes a strength rather than a weakness, and if analyzed in terms of the 
logic of analysis of variance, would result in a smaller error term rather than a 
larger one. 


In the first part of this paper, the role of anthropological data in 
editing among the competing theories of psychology has been em- 
phasized. Such research can never be central among the anthro- 
pologist's tasks, but can be invaluable in the consolidation of 
psychological theory. Anthropology is in this fashion of great meth- 
odological importance to psychology. 

In the second part of the paper, the roles are reversed. Since an- 
thropology depends upon enculturated human beings as its measur- 
ing instruments, the psychology of bias in human judgment becomes 
relevant to choices among methodological alternatives open to an- 
thropologists. Several such points are discussed, as are methodologi- 
cal strengths and weaknesses of the statistical cross-cultural studies. 


Bakan, David 

1958 Sigmund Freud and the Jewish mystical tradition. Princeton, Van 
Barry, Herbert 

1957 Relationships between child training and the pictorial arts. Journal of 
Abnormal and Social Psychology 54:380-383. 
Barry, Herbert, Margaret K. Bacon and Irvin L. Child 

1957 A cross-cultural survey of some sex differences in socialization. Journal 
of Abnormal and Social Psychology 55:3 27—3 3 2 . 


Barry, Herbert, Irvin L. Child and Margaret K. Bacon 

1959 Relation of child training to subsistence economy. American Anthro- 
pologist 61:51-63. 
Bennett, John W. 

1946 The interpretation of pueblo culture. Southwestern Journal of Anthro- 
pology 2:361-374. 

Campbell, Donald T. 

1955 The informant in quantitative research. American Journal of Sociology 

1957 Factors relevant to the vaHdity of experiments in social settings. Psycho- 
logical Bulletin 54:297-312. 

1958a Common fate, similarity, and other indices of the status of aggregates of 
persons as social entities. Behavioral Science 3:14-25. 

1958b Systematic error on the part of human Hnks in communication systems. 
Information and Control 1:334-369. 

1959 Methodological suggestions from a comparative psychology of knowl- 
edge processes. Inquiry (Univ. Oslo Press) 2:152-182. 

1 96 1 Social attitudes and other acquired behavioral dispositions. In Psy- 
chology: a study of a science. Vol. 6. Investigations of man as socius: 
their place in psychology and the social sciences. Sigmund Koch ed. New 
York, McGraw Hill (in press; dupHcated draft 1957). 
Campbell, Donald T. and Donald W. Fiske 

1959 Convergent and discriminant vaHdation by the multitrait-multimethod 
matrix. Psychological Bulletin 56:81-105. 
Campbell, Donald T., William A. Hunt and Nan A. Lewis 

1957 The effects of assimilation and contrast in judgments of cHnical ma- 
terials. American Journal of Psychology 70:347-360. 

1958 The relative susceptibiHty of two rating scales to disturbances resulting 
from shifts in stimulus context. Journal of Applied Psychology 42: 

Chapman, Loren J. and R. Darrell Bock 

1958 Components of variance due to acquiescence and content in the F scale 
measure of authoritarianism. Psychological Bulletin 55:328-333. 
Child, Irvin L. 

1954 SociaUzation. In Handbook of social psychology, Gardner Lindzey, ed. 
Cambridge, Mass., Addison-Wesley. 
Child, Irvin L., Thomas Storm and Joseph Veroff 

1958 Achievement themes in folk tales related to sociahzation practice. In 
Motives in fantasy, action, and society. John W. Atkinson, ed. New 
York, Van Nostrand. 
Cochran, William G. and Gertrude M. Cox 

1950 Experimental designs. New York, Wiley. 
Cronbach, Lee J. 

1946 Response sets and test vaHdity. Educational and Psychological Measure- 
ment 6:475—494. 

1950 Further evidence on response sets and test design. Educational and 
Psychological Measurement 10:3-31. 


DooB, Leonard W. 

1957 An introduction to the psychology of acculturation. Journal of Social 
Psychology 45 : 143-1 60. 

1958 The use of different test items in nonUterate societies. Public Opinion 
Quarterly 21:499-504. 

Eggan, Dorothy 

1952 The manifest content of dreams: a challenge to social science. Ameri- 
can Anthropologist 54:469-485. 
Freeman, Linton C. and Robert F. "Winch 

1957 Societal complexity: an empirical test of a typology of societies. Ameri- 
can Journal of Sociology 62:461-466. 
Gladwin, Thomas 

1954 Review of "Whiting and Child, Child training and personality. American 
Anthropologist 56:893-897. 
GiLLEN, John 

1954 Methods of approach to the study of human behavior. In Aspects of 
culture and personality, Francis L. K. Hsu, ed. New York, Abelard- 
Heider, Fritz 

1944 Social perception and phenomenal causality. Psychological Review 51: 

Helson, Harry 

1947 Adaptation-level as frame of reference for prediction of psycho- 
physical data. American Journal of Psychology 60:1—29. 
Henry, Jules and Zunia 

1944 Doll play of Pilage Indian children. Research Monographs No. 4, Ameri- 
can Orthopsychiatric Association. 
Herskovits, Melville J. 

i960 Humanism in anthropological science. Yl International Congress of 
Anthropological and Ethnological Sciences, Paris, August 4, i960. 
Herskovits, Melville J. and Frances S. 

1958a Sibling rivalry, the Oedipus complex, and myth. Journal of American 

Folklore 71:1-15. 
1958b Dahomean narrative: a cross-cultural analysis. Evanston, 111., North- 
western University Press. 
Holmes, Lowell Don 

1957 The restudy of Manu'an culture: a problem in methodology. Unpub- 
lished Ph.D. dissertation, Northwestern University. 

1958 Ta'u: Stability and change in a Samoan village, "Wellington, New Zea- 
land, Polynesian Society, Reprint No. 7. 

Honigmann, John J. 

1952 The testing of hypotheses in anthropology. American Anthropologist 

1954 Culture and personality. New York, Harper. 

HoRTON, Donald 

1943 The functions of alcohol in primitive societies: a cross-cultural study. 
Quarterly Journal of Studies on Alcohol 4: 199-320. 


Hsu, Francis L. K. 

1940 The English wife. Chapter VI, Mother and children, Section 5, A non- 
Freudian explanation. Unpublished manuscript. 
1954 Aspects of culture and personality. New York, Abelard-Schuman. 
Keller, Albert Galloway 

193 I Societal evolution (rev. ed.), New Haven, Yale University Press. 
Kish, Leslie 

1956 Confidence intervals for clustered samples. American Sociological Re- 
view 22:154-165. 
Kluckhohn, Clyde 

1954a Culture and behavior. In Handbook of social psychology, Gardner 

Lindzey, ed. Cambridge, Mass. Addison-Wesley. 
1954b Southwestern studies of culture and personaHty. American Anthro- 
pologist 56:685-697. 

Lambert, "William "W., Leigh Minturn Triandis and Margery Wolf 
1959 Some correlates of beliefs in the malevolence and benevolence of super- 
natural beings: a cross-cultural study. Journal of Abnormal and Social 
Psychology 58:162-169. 
Lasswell, Harold D. 

193 I A hypothesis rooted in the preconceptions of a single civilization tested 
by Bronislaw Malinowski. In Methods in social science, Stuart A. Rice, 
ed. Chicago, University of Chicago Press. 
Levy, David M. 

1935 Sibling rivalry studies in children of primitive groups. American 
Journal of Orthopsychiatry 9:205—214. 
Lewis, Oscar 

195 1 Life in a Mexican village: Tepoztlan restudied. Urbana, University 
of Illinois Press. 

Li An-Che 

1937 Zuni: some observations and queries. American Anthropologist 39: 
Malinowski, Bronislaw 

1927 Sex and repression in savage society. London, Humanities Press. 
McClelland, David C. and G. A. Friedman 

1952 A cross-cultural study of the relationship between child-training prac- 
tices and achievement motivation appearing in folk tales. In Readings 
in social psychology, Guy E, Swanson, Theodore M. Newcomb, and 
Eugene L. Hartley, eds. New York, Holt. 

Mead, Margaret and F. M. C. MacGregor 

195 1 Growth and culture: a photographic study of Balinese childhood. New 
York, Putnam. 


1946 La perception de la causalite. Louvain, Institute superior de Philosophic, 
Etudes Psychologiques vol. 6. 

MuRDOcK, George Peter, and John "W. M. Whiting 

195 1 Cultural determination of parental attitudes: the relationship between 


the social structure, particularly family structure, and parental be- 
havior. In Problems of infancy and childhood. Transactions of the 
Fourth Conference March 6—y, 1950, Milton J. E. Senn, ed. New York, 
Josiah Macy Jr. Foundation, 195 1. 
Murray, Henry A. 

1949 Research planning: a few proposals. In Culture and Personality, S. 
Stansfeld Sargent and Marian W. Smith, eds. New York, Viking Fund, 

Paul, Benjamin D. 

1950 Symbolic sibling rivalry in a Guatemalan Indian village. American 
Anthropologist 52:205—217. 

Pettitt, George A. 

1946 Primitive education in North America. Berkeley, University of Cali- 
fornia Press, University of California Publications in American Archae- 
ology and Ethnology Vol. 43, No. i. 
Rose, Edward and Gary Willoughby 

1958 Culture profiles and emphases. American Journal of Sociology 63: 
Sargent, S. Stansfeld and Marian W. Smith 

1949 Culture and Personality. New York, Viking Fund. 
Spiro, Melford E. 

1953 Ghosts: an anthropological inquiry into learning and perception. Journal 
of Abnormal and Social Psychology 48:376—382. 

Spiro, Melford D. and Roy G. D'Andrade 

1958 A cross-cultural study of some supernatural beliefs. American Anthro- 
pologist 60:456—466. 

Wellisch, E. 

1954 Isaac and Oedipus: a study in Biblical psychology of the sacrifice of 
Isaac, the Akedah. London, Routledge & Kegan Paul. 

Whiting, Beatrice B. 

1950 A cross-cultural study of sorcery and social control. In Paiute Sorcery. 
New York, Viking Fund (Publications in Anthropology No. 15). 

Whiting, John W. M. 

1954 The cross-cultural method. In Handbook of social psychology, Gardner 
Lindzey, ed. Cambridge. Mass., Addison-Wesley. 

1959 Sorcery, sin and the superego: a cross-cultural study of some mecha- 
nisms of social control. In Nebraska symposium on motivation: 1959, 
Marshall R. Jones, ed., pp 174-195. Lincoln, University of Nebraska 
Press, 1959. 

Whiting, John W. M. and Irvin L. Child 

1953 Child training and personality. New Haven, Yale University Press. 
Whiting, John W. M., Richard Kluckhohn and Albert Anthony 

1958 The function of male initiation ceremonies at puberty. In Readings in 
social psychology, 3d ed,, Eleanor E. Maccoby, Theodore M. Newcomb, 
and Eugene L. Hartley, eds. New York, Holt. 
Wright, George O. 

1954 Projection and displacement: a cross-cultural study of folk-tale ag- 
gression. Journal of Abnormal and Social Psychology 49:523-529. 


If psychological characteristics of the individual, whether iden- 
tified with his total personality or with the socially functioning part 
of it, are dependent upon the culturally conditioned child-rearing 
practices or socialization processes, what are the factors which de- 
termine or at least shape the patterns of culture, which in turn 
condition the child-rearing practices or socialization processes? If 
human societies are as stable and unchanging as those of ants and 
bees, the latter type of question, though not wholly irrelevant, 
would not have been important. But human societies are highly 
dynamic entities with extreme variability in their rates of change, 
just as human individuals in any society are quite capable of, and 
often given to, deviation. 

Jules Henry, a well-known psychological anthropologist, puts it 
this way: "As I see it, the crucial difference between insect societies 
and human ones is that whereas the former are organized to achieve 
homeostasis, the organization of the latter seems always to guarantee 
and specifically provide for instability" ("Homeostasis, Society 
and Evolution: A Critique." Scientific Monthly, LXXXI, 1955: 
308). While this may be an overstatement, the plain fact is that 
all human societies do undergo change, rapidly or slowly. 

The question of individual deviation was discussed by Kaplan in 
Chapter 8; the question of social and cultural change was briefly 
touched upon by Hsu in Chapter 7. The best accepted view at pres- 
ent is that the individual and society-culture relationship is a two- 
way traffic in spiral progression. The individual's psychological 
characteristics are results of his socialization processes, but his psy- 
chological characteristics are, in turn, at the root of the patterns of 
culture, in change or in stability, which govern the socialization 

The three chapters in this section of the book have some impor- 



tant differences which the reader will do well to keep in mind. The 
first difference concerns approach. Whiting's approach and that of 
Aberle are more rigorous in methodology, with emphasis on ascer- 
taining cross-culturally the interrelationship between a few specific 
variables (such as "exclusive mother-infant sleeping arrangement" 
and "cross sex identity," or "economic organization" and "ethics") . 
Hsu's approach, while likewise attempting cross-cultural generali- 
zation, is still at a more qualitative or speculative stage. The gen- 
eralization attempted is perhaps for this reason more "ambitious," 
in that it hypothesizes the existence of a single socio-psychological 
axis that generates or integrates a wide range of more specific cul- 
tural features. While the Whiting and Aberle chapters in this sec- 
tion, as well as the other chapters of the entire book, are primarily 
critical appraisals of works already carried out or well under way, 
Hsu's chapter is launched more or less as a trial balloon, an explora- 
tion of a hypothesis which will stand or fall depending upon in- 
tensive research yet to come. 

chapter 12 


Harvard University 

The use of the comparative or cross-cultural method in studies of 
culture and personality has served two quite different purposes. 
Psychologists have tended to view this method as one by which cer- 
tain assumptions about personality development may be tested. 
Anthropologists, on the other hand, are more likely to view such 
studies as a test of hypotheses concerning the way in which elements 
of culture can be integrated by underlying psychological processes. 
It is to the latter aim that this chapter will be devoted. 

Most of the early studies concerned with culture and personality 
were intensive case studies of a single society such as Mead's Coining 
of Age in Samoa (1928) or comparisons of a series of case studies 
such as Ruth Benedict's Vat terns of Culture (1934), Margaret 
Mead's Sex and Temperament (1935) , Linton and Kardiner's The 
Individual and His Society (1939) and The Psychological Fron- 
tiers of Society (1945). This review, however, will not consider 
such case studies, but will be restricted to cross-cultural studies 
which have used a large sample of societies presumed to be in some 
way representative of the cultures of the world. 

The studies under review can be classified essentially into two 
types: those which have made some assumptions about the psycho- 
logical effect of certain child-rearing practices on personality as re- 
flected in some other aspect of culture such as magic, art, or religion; 
and those which have concerned themselves with the effect of fea- 
tures of the basic economy or social structure on child-rearing prac- 
tices. Fortunately, in many instances studies in these two categories 
may be linked by virtue of the fact that they share the same scores 



on child-rearing practices. In the first type of study these Unking 
child-rearing scores have the theoretical status of independent or 
antecedent variables; that is, they have been assumed to be deter- 
minants of personality which is assumed to be a mediating psycho- 
logical process reflected in magic and religion. In the second type 
of study, child-rearing scores have the theoretical status of de- 
pendent or consequent variables; that is, they have been assumed to 
be determined by economic and social structural aspects of the cul- 

The conjunction of these two kinds of studies described above 
permits the testing of the general hypothesis suggested by Whiting 
and Child (1953:310) concerning the way in which personality or 
psychological process may serve to integrate culture. This hypothe- 
sis was summarized by the following diagram: 

Maintenance Child Training Personality Projective 

Systems Practices Variables Systems 

Maintenance systems were defined as ''the economic, political, 
and social organizations of a society — the basic customs surrounding 
the nourishment, sheltering, and protection of its members." Per- 
sonality was defined as "a set of hypothetical intervening variables." 
Projective systems include customs which are for the most part 
magical and unrealistic. The term "projective system" suggested 
by Kardiner (1945) is perhaps unfortunate since it suggests that 
the psychological process of projection is necessarily involved. Since 
"acting out," "distortion," "ritualization," "displacement," "fixa- 
tion," or any other psychological process relating to personality 
is implied, a term such as "systems of psychological defense" or of 
"psychological security" might have been more appropriate. The 
cultural systems which reflect such processes most directly are those 
of magic, religion, art or any other feature that is not immediately 
and practically involved in the satisfaction of basic biological needs. 
In sum, the hypothesis implies that personality is an intervening 
hypothetical variable determined by child rearing which is in turn 
determined by maintenance systems and which finally is reflected 
in projective systems. 

This paper, then, will review the evidence for and against this 
general hypothesis. The evidence will be drawn from cross-cultural 
studies of the two types specified above. This review could be or- 
ganized by maintenance systems, child-rearing variables, interven- 
ing psychological processes, or projective variables. I have rather 


arbitrarily chosen to organize it by child-rearing practices which 
have been ordered in terms of the life line from early infancy to 
later childhood. I will begin, therefore, with those studies related 
to the treatment of infants. 

Parental Image and the Nature of the Gods 

For a long time psychologists, particularly those of Freudian per- 
suasion, have assumed that the nature of the gods and their relation 
to man is a reflection of the parental image and, hence, could be 
predicted from the relation between parent and child during in- 
fancy and early childhood. Several cross-cultural studies have re- 
cently attempted to put this hypothesis to the test (Spiro and 
D'Andrade 1958; Lambert, Triandis, and Wolf 1959; and Whiting 
1959a) . Each of these studies tends to support the general hypothe- 
sis that harsh parental treatment during infancy leads to the cul- 
tural belief that the spirit world is harsh and aggressive. 

Spiro and D'Andrade (1958), using the Whiting and Child 
(1953) "initial satisfaction of dependence" as a score ^ for esti- 
mating the degree to which infants are indulged, found that socie- 
ties that were judged to be relatively high on the above score tended 
to believe that the behavior of the gods was contingent upon the 
behavior of humans and that gods could be controlled by the per- 
formance of compulsive rituals.^ Such societies did nof propitiate 
the gods. The authors argue that the adults' treatment of the gods 
is, therefore, a reflection of an infant's relation to his parents. In 
other words, infants who are treated indulgently by their parents, 
that is, whose parents respond to them when they cry or show dis- 
comfort, when they grow up feel they can be equally successful in 
controlling the supernaturals. 

Lambert, Triandis, and Wolf (1959) used a score taken from 
Barry, Bacon, and Child (1957) for estimating the relation be- 
tween an infant and his caretakers, consisting of a judgment of the 
degree to which they treated him harshly or painfully. They found 
that societies in which infants were treated relatively painfully be- 
lieved in gods which were judged to be more aggressive than benevo- 
lent toward human beings. Again the gods seem to reflect the par- 
ental treatment of infants. 

This score includes such items as the encouragement of the infant's dependence, his freedom 
to be dependent, and the duration of this freedom. For a more complete description of this 
score see Whiting and Child (1953), pp. 50, 91. 

^ Unless specified the 5 per cent level of confidence or better has been used as a criterion to 
report a relationship. To simplify presentation p values will not ordinarily be reported. 


Finally, Whiting (1959a) , using still a different score for infant 
indulgence, reports a finding consistent with this hypothesis. The 
score in this study was also from Barry, Bacon, and Child (1957) 
and was an over-all judgment of the degree to which an infant was 
indulged by his caretakers.^ It was reported that societies high in 
the over- all indulgence of infants tended not to fear ghosts at 
funerals. The assumption here is that funereal ghosts are, like the 
gods in the previous studies, a projection of the parental image. 

In order to test the general hypothesis of personality as a medi- 
ator, the next problem is to discover whether or not there is any 
relationship between maintenance systems of a culture and the de- 
gree to which infants are indulged. It was suggested by Murdock 
and Whiting (1951) that the economic and ceremonial duties of 
the mother might have some bearing on the amount of time she 
could spend in caring for her child, and tentative results based on 
a small number of cases tended to confirm this hypothesis. They 
report (pp. 33—35) that societies in which mothers have few eco- 
nomic responsibilities and are little involved in the ceremonial life 
of the tribe tend to be more indulgent with their infants than in 
societies where mothers have such responsibilities. These results 
were based on a very small sample of societies and were not statis- 
tically significant and, therefore, must be judged as highly tenta- 
tive. They also reported that there was a tendency for large extended 
families where there were many hands to care for the infant, to treat 
him more indulgently. Again this relationship was not strong and 
reached only the 10 per cent level of statistical significance. Mur- 
dock (1957), however, has recently published judgments on the 
family and household structure for a large number of societies. 
This, taken together with the ratings by Barry, Bacon, and Child 
(1957) on the degree of over-all indulgence described above, en- 
ables us to make a more adequate test of this hypothesis than was 
possible in 195 1. Since household membership rather than family 
structure should be most relevant to our hypothesis, this has been 
used as our independent variable. The results of the test are pre- 
sented in Table i . 

It will be seen from this table that the degree of infant indul- 
gence is roughly proportional to the number of adults living in the 
household. Extended and polygynous families where there are more 

^This score took account of the following items: display of affection, degree of drive reduction, 
immediacy of drive reduction, constancy of the presence of caretakers, and the absence of pain 
induced by caretakers. 






























Ontong Java 


















Lesu (11) 







Kurtachi (12) 







Bena (13) 







Chukchee (11) 







Ainu ( 5 ) 










Ashanti (10) 
Azande (10) 
Chagga (7) 
Dahomeans (7) 




Ganda (9) 

Pukapukans (9) 

Masai (10) 

Mbundu (9) 

Tanala (9) 

Thonga (7) 

Venda (9) 

W.Apache (10) 

Table i. The relation between household structure and the over-all indulgence of infants. 
The numbers following the names of the societies indicates the Barry, Bacon, and Child (1957) 
score on over-all infant indulgence. Extended households include lineal and stem as well as large 
extended categories of Murdock. The two communal households in the sample — Siriono (10) and 
Yagua (11) — are omitted from the table. 

than two adults living in the household tend to be predominantly 
indulgent with their infants. Nuclear households with two adults 
are unpredictable. Finally, in the mother-child household where one 
woman alone has to care for her children the probability of high in- 
dulgence is slight. The percentage of societies with high infant in- 
dulgence is as follows: extended, 87 per cent; polygynous, 83 per 
cent; nuclear, 42 per cent; and mother-child, 25 per cent. The 
probability that both extended and polygynous households will be 
high on infant indulgence is statistically significant at better than 
the 5 per cent level of confidence. Societies with nuclear house- 
holds are unpredictable in this respect. Although only 25 per cent 
of societies with mother-child households are indulgent, this rela- 
tionship does not quite reach an acceptable level of confidence. 


Before we can accept the thesis that infant indulgence creates a 
parental image which is reflected in the gods and thus forms a link 
between household structure and religious beliefs, we must meet 
the argument that household structure and the nature of the gods 
are related to one another for some other reason and that they 
jointly affect the treatment of children. If this latter hypothesis 
were true, the gods could be predicted from a knowledge of the 
household structure when the child-rearing factors were held con- 
stant. This is not, in fact, the case. Thus, for example, the Tenete- 
hara who, although they have an extended household, are excep- 
tional in being rated low in the indulgence of infants, have 
aggressive gods, a fact which would have been predicted from their 
child rearing rather than from their household arrangements. Con- 
versely, the Chukchee who, although they have mother-child 
households, are high in the indulgence of their children — an excep- 
tion to the rule that mother-child households are low in infant in- 
dulgence — have benevolent gods. 

Thus, child rearing rather than household structure seems to be 
the determinant of the nature of the gods. Statistically, household 
structure can be shown to be unrelated to the gods if infant indul- 
gence is not taken into account.^ Thus, although 87 per cent of ex- 
tended family households are high on infant indulgence and 80 per 
cent of the societies with high indulgence are below average on the 
fear of ghosts at funerals, only 6y per cent of the extended families 
in the samprle are below average on fear of ghosts. The relation be- 
tween household and indulgence and that between indulgence and 
ghost fear are statistically significant at better than the i per cent 
level of confidence. Thus, it seems that the nature of the gods can- 
not be predicted from a knowledge of household structure alone. 
Child rearing with its influence on personality seems to be prerequi- 

Exclusive Sleeping Arrangements and Cross Sex Identify 

The over- all indulgence of infants discussed above is concerned 
with how a child is treated during the day. The relation of a child 
to his parents at night has also been shown (Whiting et al. 1958) 
to be an important child-rearing variable. In most societies over 
the world infants sleep in the same bed or on the same sleeping mat 

^This methed of analysis is similar to that suggested by Blalock (i960). Confidence limits 
(Hald 1952) rather than correlation coefficients have been used to establish the relative degree 
relationship between the three variables. 


with their mothers. Even where an infant has a cradle or cot of his 
own, this is generally placed next to the mother's bed within easy 
reach. The sleeping distance between a mother with a nursing in- 
fant and her husband, however, is more varied. In slightly over 
half of the societies of the world the husband sleeps either in a bed 
in the same room but at some distance from his wife, or in another 
room. This may be called an "exclusive mother-infant sleeping ar- 

Whiting and co-workers (1958) showed that exclusive mother- 
infant sleeping arrangements are strongly associated with male initi- 
ation rites at puberty. They offered three different interpretations 
of this association. They assumed that such sleeping arrangements 
( I ) increased the Oedipal rivalry between son and father and that 
initiation rites served to prevent open and violent revolt against 
parental authority at a time when physical maturity would make 
such revolt dangerous and socially disruptive, (2) lead to exces- 
sively strong dependence upon the mother which initiation rites 
serve to break, and (3) produced strong identification with the 
mother which the rites serve to counteract. 

Although the first interpretation was favored by these authors, 
later research (Whiting 1960a; Burton and Whiting i960; 
Stephens, ms.) has favored either the third or a modification of the 
second, the incest hypothesis to be discussed below. The first inter- 
pretation has been rejected for a number of reasons. The assump- 
tion made by Whiting and his associates (1958) that exclusive 
mother-infant sleeping arrangements exacerbate rivalry between 
father and son is not supported if one looks more closely at the 
facts. In the first place, since such sleeping arrangements usually 
occur in polygynous societies, the father has sexual access to his 
other wife and, hence, should not be particularly frustrated by the 
infant or see him as a rival. In the second place, at the time of wean- 
ing when the exclusive sleeping arrangements terminate, the father 
usually does not move in to sleep with the mother, since in more 
than half such societies a man never sleeps with his wife and in most 
of the remaining societies he sleeps with each wife in turn and, 
thus, sleeps with any one wife at most but half the time. 

Campbell has in this volume suggested another version of the 
rivalry hypothesis, namely, that a younger sibling may be seen as 
the person responsible for the infant's fall from grace at the time 
of weaning. Although this hypothesis has considerable plausibility, 
the fact that in societies with exclusive mother-infant sleeping 


arrangements the mother is under a sex taboo during the nursing 
period should mean that the younger sibhng would ordinarily not 
appear until at least nine months after the previous child's dis- 
placement. The mother, herself, therefore, seems to be the best 
candidate as the person who is perceived by the child as the one 
responsible for the termination of his exclusive relationship with 
her. It is she who at the same time both weans him and refuses to let 
him sleep with her. 

In a recent theoretical paper Whiting (1960b) has formulated a 
series of hypotheses concerning identification as it relates to the con- 
trol and mediation of resources. One hypothesis in this formulation 
has bearing upon the analysis in the preceding paragraph. This, the 
so-called "status-envy hypothesis," is stated by Whiting (1960b: 
18) as follows: "If a child perceives that another has more efficient 
control over resources than he has; if, for example, he sees another 
person enjoying resources of high value to him when he is deprived 
of them, he will envy such a person and attempt to emulate him." 

If the status-envy hypothesis be applied to sleeping arrange- 
ments, the father should be seen to occupy an envied position if he 
sleeps with the mother, particularly if the infant is in a cradle. 
Contrariwise with the exclusive mother-infant arrangements, when 
the mother withdraws this exclusive privilege at the time of wean- 
ing, she should be seen as the most envied person. This should lead a 
boy to see his mother's status, and that of women in general, as 
being all important and powerful, and, hence, lead to cross sex 

A preliminary test of this hypothesis was presented by Whiting 
(1960a) and has been summarized by Burton and Whiting (i960). 
A more detailed report is in preparation and will be published un- 
der the joint authorship of Whiting, Fischer, D'Andrade, and 
Munroe. In this study the following evidence is presented in sup- 
port of the status-envy hypothesis. 

First, members of the societies in which male initiation rites oc- 
cur often define these rites as death and rebirth — the death of a 
person in a "woman-child" status and rebirth into the status of an 
"adult male." This suggests that an initial cross sex identification in 
boys is recognized. 

Second, exclusive mother-infant sleeping arrangements are as- 
sociated with the couvade as well as male initiation rites. The 
couvade can be interpreted as a cultural device which permits the 
acting out of the female role. Since initiation rites and couvade 


rarely occur in the same society, some reason must account for the 
choice between counteracting and permitting the expression of 
cross sex identity. Residence patterns serve this purpose. Societies 
with exclusive mother-infant sleeping arrangements and patrilocal 
residence tend to have initiation rites, whereas those with exclusive 
sleeping and matrilocal residence generally have the couvade. It has 
not been settled as to whether residence operates as another factor 
relating to status envy and identification or whether it requires a 
differential role for adult males. 

Third, totemism was also shown to be associated with exclusive 
mother-infant sleeping arrangements. This fact leads to the inter- 
pretation that totemism serves to establish a male's relationship to 
his male progenitors where his early life creates some doubt about it. 

Finally, in a recent study Bacon, Child, and Barry (ms.) showed 
that the rate of personal crime (assault, murder, rape, suicide, 
sorcery, and the making of false accusation) is highest in societies 
with exclusive mother-infant sleeping arrangements. They in- 
terpreted this as an attempt, in part at least, to express masculinity 
in societies where there is a need to deny an underlying feminine 

As has already been suggested, polygyny is the maintenance sys- 
tem variable most strongly associated with exclusive sleeping ar- 
rangements. In nearly 80 per cent of societies with strict monogamy 
the mother and father sleep in the same or adjacent beds, whereas 
this is only true of 3 per cent of those households where a husband 
has more than one wife. Whether polygyny has an influence upon 
the various projective consequences of exclusive mother-infant 
sleeping arrangements is now under investigation and cannot be 
reported upon here. Residence, however, as was reported above does, 
in interaction with sleeping arrangements, have a direct association 
with both male initiation rites and the couvade. 

Infant Seduction and Mother-Son Incest 

Whiting and his co-workers (1958) found that another child- 
rearing practice relating to infancy was strongly associated with 
male initiation rites at puberty. This practice consists of a prolonged 
postpartum sex taboo lasting for at least a year. This practice is 
often associated with the belief that sexual intercourse will sour or 
alter the mother's milk in a manner that would be dangerous to a 
nursing infant. The taboo is generally coterminous with the nurs- 
ing period which often lasts in these societies for nearly three years. 


Whiting's group (1958) interpreted this factor as having much the 
same effect as that of an exclusive mother-infant sleeping arrange- 
ment. Stephens (ms.), however, assumed that a mother, deprived 
of her normal sex life during such a prolonged period, will gain some 
indirect sexual satisfaction from her infant, particularly during the 
act of nursing. If this interpretation is correct, a strong incestuous 
bond between mother and son should be established in societies with 
a prolonged postpartum sex taboo. 

Stephens (ms.) argued that since the expression of mother-son 
incest is not permitted in adult life in any society, this early tend- 
ency must be strongly opposed and that strong sex conflict and 
anxiety should be induced. As a projective index of such conflict, 
he chose the degree to which menstrual taboos were elaborated in a 
society. He established a scale which indicated the degree to which 
women were isolated from men while they were menstruating and 
argued that this measured castration anxiety in the males. He then 
showed that societies with a prolonged postpartum sex taboo tended 
to have elaborate menstrual taboos as measured by this scale. The 
fact that the Whiting and Child ( 1953 ) measure of the severity of 
sex training in later childhood was also related to Stephen's men- 
struation scale lends support to the interpretation that it is an in- 
dicator of sex anxiety. 

These results suggest that male initiation rites serve to oppose 
mother-son incest as well as to counteract cross sex identification. 
The fact that severe menstrual taboos were not found by Stephens 
to be independently related to male initiation rites is puzzling, how- 

Stephens and D'Andrade (Stephens, ms.) report still another 
consequence of a prolonged postpartum sex taboo. They showed 
that societies with this practice tend to have formal avoidance pat- 
terns between a woman and her daughter's husband, between a man 
and his son's wife, and between a brother and a sister. They argue 
that these avoidances result from sexual conflict produced by the 
seductive and incestuous relationship between mother and infant 
consequent upon the prolonged postpartum sex taboo. 

Polygyny is again the aspect of the maintenance system which is 
highly predictive of a prolonged postpartum sex taboo. Stephens 
(ms.) , however, found that polygyny alone is not significantly re- 
lated to the degree of elaboration of menstrual taboos. Thus, a pat- 
tern similar to that reported for the nature of the gods emerges, 
where a maintenance system variable is related to a projective sys- 


tern variable by common linkage with a personality variable implied 
by a child-rearing practice. 

The Age of Sodolizotlon— Guilt 

Proceeding along the life line of the child, the next item that has 
been used in cross-cultural research concerns variations in the age 
at which societies begin the serious training of their children. This 
has in general been shown to affect the projective systems which 
reflect guilt. Whiting and Child (1953), taking as a measure of 
guilt the degree to which a patient was believed to be responsible 
for causing his own illness, presumably indicating his readiness to 
accept blame, found that societies with early weaning, early inde- 
pendence training, and early training in modesty and the inhibition 
of heterosexual play were those which tended to have high guilt. 
The age of toilet training was not related. 

Whiting and Child (1953) tentatively concluded that this rela- 
tionship was due to identification. Anticipating the status-envy 
hypothesis, they argued that parents should seem more powerful 
to a very young child than to an older one who has already learned, 
to a degree at least, to cope with the environment by himself. Thus, 
early socialization should produce stronger identification and, 
hence, guilt over contravening parental values. 

It is again possible to relate this association to the maintenance 
systems. Whiting (1959b) reports that household structure is a 
significant determinant of the age of sociahzation. Nuclear house- 
holds are earliest for both weaning (median age 2 years) and inde- 
pendence training (median age 2 years, 9 months) and mother- 
child households are the latest.^ On the average they do not begin 
to wean their children until they are three years old nor start train- 
ing them in independence until they are four and one half. Extended 
and polygynous households fall in between these two extremes for 
both weaning and independence training. 

To test our hypothesis we have to ask whether nuclear households 
independently of child rearing have higher guilt than mother-child 
households. This is in fact what is reported by Whiting (1959a) ; 
86 per cent of the nuclear households in the sample reported had 
high scores on patient responsibility whereas but 14 per cent of the 
mother-child households were high in this regard. 

Although it is difficult to be sure with the relatively small num- 

' The age of independence training should not be confused with the degree of infant indulgence 
referred to above (p. 358). Mother-child households are both low and late. 


ber of cases on which data are available, it seems that in this instance 
households have some effect on guilt independent of child rearing. 
Thus, whereas it seems as if the nature of the gods is directly de- 
pendent on child-rearing practices and only indirectly upon house- 
hold structure, guilt is produced by an interaction of both social 
structure and child rearing. Thus, Whiting (1959a) showed that 
the age of weaning was correlated with patient responsibility for 
monogamous societies. This was not true of polygynous societies. 
They tended to have a low score on guilt whether weaning was 
early or late. From this we may conclude that, while the age of so- 
cialization may be a mediating factor between social structure and 
magical theories of disease, it is clearly not the only one. 

It was assumed above that strong identification with the mother 
should be induced by exclusive mother-infant sleeping arrange- 
ments which is in turn strongly associated with polygyny. But 
polygynous societies are low in guilt. To get around this contradic- 
tion. Whiting (1959c) has argued that guilt is derived from identi- 
fication with the male rather than the female role. The basis of 
argument consisted of the assumption that "the role of the father 
and of males in general in any society tends to be more punitive, 
rigid and unforgiving than that of the mother and of women in 
general .... A woman could scarcely bring up a child unless, when 
he deviates from the familial rules, she made exceptions if the child 
were sick or tired or upset." 

Severity of Socialization and Negative Fixation 

Estimates of the severity of socialization in early childhood pro- 
vide the next set of child-rearing variables to be considered. Such 
estimates were made by Whiting and Child (1953) with respect to 
five systems of behavior: oral, anal, sexual, aggression, and de- 
pendence. The presumed effect of severe training was that of "nega- 
tive fixation" or the anxious preoccupation with the type of be- 
havior or behavior system which is severely punished. The theory 
of negative fixation was based upon the effect of conflict rather 
than on the stages of psychosexual development. The hypothesis 
which they put forward is that conflict between habits learned in 
infancy and then punished during the socialization process pro- 
duces persistent motivation which activates behavior in adulthood 
in some way related to the conflict and presumably is functionally 
defensive in nature. 

Explanations for illness and therapeutic techniques were chosen 
by Whiting and Child (1953) as aspects of the projective system 


which might reflect fixation. A content analysis of magical beliefs 
and practices relating to illness was made for each society with the 
five behavior systems in mind. In judging the severity of socializa- 
tion for each system, the following factors were taken into con- 
sideration: intensity and frequency of punishment, suddenness of 
the transition from behavior appropriate to infancy and that to 
later childhood, and signs of emotional disturbance on the part of 
the child.^ 

In general the fixation hypothesis was supported. The severity of 
weaning (oral anxiety) was strongly related to "oral explanations 
for illness." Such oral explanations include the belief that sickness 
is caused by eating or drinking magically poisoned food or by the 
verbal spells and incantations of sorcerers. The severity of aggres- 
sion training (aggression socialization anxiety) which includes the 
treatment of temper tantrums, physical and verbal aggression, 
damage to property, and disobedience was related to explanations 
for illness involving aggression. These include hostility toward or 
disobedience to spirits, poison if it is introjected into the patient 
rather than being ingested, and the use of magical weapons by a 
sorcerer. The severity of independence training was shown to be 
related to dependence explanations for illness, a measure which in- 
cludes the belief that illness could be caused by "soul stealing" or 
by "spirit possession." The negative fixation hypothesis was not 
confirmed in the other two systems of behavior. Toilet training did 
not predict the Whiting and Child score on anal explanations for 
illness nor did the severity of sex training predict sexual explana- 
tions for illness. However, there was some indication that relevant 
avoidance in these behavior systems was used as a therapeutic prac- 
tice. Thus, societies with severe toilet training tend to have thera- 
peutic practices involving washing or cleansing, the adherence to 
cleanliness taboos, or the retention of feces, and societies with severe 
sex training tended to believe that abstention from sexual inter- 
course by the patient would have a therapeutic effect. 

In addition, as was reported above, Stephens (ms.) found that 
severe sex training is associated with elaborate menstrual taboos, and 
Ayres (1954) showed this child-rearing measure to be related to 
prolonged sex taboos during pregnancy. Each of these may be 
viewed as an index of negative fixation. 

The following maintenance system variables have been reported 

It should be noted that the age of socialization was conceptually distinguished from the 
severity of socialization. Although in general these measures were negatively correlated (Whiting 
and Child 1953, p. no), they were empirically distinct as well. In other words late socialization 
is not necessarily mild. 


to be associated with severity of socialization in the various systems. 
Murdock and Whiting (1951) report that societies with sororal 
polygyny are significantly less severe in weaning their children than 
are societies with nonsororal polygyny. Monogamous societies, ac- 
cording to their findings, stand between these two extremes and are 
not significantly different from either. They explained mild wean- 
ing in sororal-polygynous societies as a consequence of the co-opera- 
tion between co-wives who are sisters. The severity of sex training 
is associated with polygyny. Only 1 5 per cent of the societies which 
are monogamous, or in which not more than 10 per cent of the 
women are polygynously married, are above the median on the 
severity of sex training, whereas 73 per cent of the societies with a 
higher proportion of polygynous marriages are severe in this regard. 

Finally, a strong association between the severity of aggression 
training and household structure has been reported by Whiting 
(1959b). Ninety-two per cent of the extended families in the 
sample used are above the median on the punishment for aggres- 
sion. Nuclear households were least severe in this respect — only 
25 per cent of the cases being severe. Polygynous and mother-child 
households were 61 per cent and 46 per cent respectively. Whiting 
(ms.) in an analysis of the Zuni extended family households sug- 
gests that the expression of aggression cannot be tolerated in cir- 
cumstances where so many people are living in such crowded quar- 
ters. No maintenance system variable has as yet been reported to 
predict the severity of either toilet training or independence train- 
ing. An item of interest, however, that should be followed up was 
reported to me by C. S. Ford. An undergraduate paper in one of his 
classes showed that toilet training was more severe in societies that 
had wooden floors and rugs than in societies with dirt floors and no 

Factor analysis provides another method of estimating the effect 
of child-rearing practices upon projective systems. Prothro (i960) 
subjected the Whiting and Child (1953) fixation hypothesis to 
such an analysis. The first factor, which he names the "aggression- 
hypochondriasis factor" had high positive loadings for the severity 
of aggression training, and all explanations for illness and tech- 
niques of therapy save those relating to dependence. This factor was 
also positively loaded on sorcery and negatively on the fear of 
spirits. The second factor, named "orality-sexuality," had heavy 
negative loadings on initial indulgence for dependent and oral sys- 


terns, and on the severity of sex training. It also had heavy positive 
loadings on oral and dependent explanations for illness and the fear 
of spirits. The third and final factor, "independence-anality," had 
a high positive loading on the severity and earliness of toilet training, 
and a strong negative loading on the severity and earliness of inde- 
pendence training. A negative loading dependence avoidance 
therapy was the only projective measure which seemed related to 
this factor. 

Seventy of Socialization: Projection and Displacement 

Cross-cultural studies involving the importance of sorcery and 
witchcraft have generally interpreted this belief as one involving 
the psychological mechanism of projection and/or displacement. 
Two views of this mechanism have been put forth. One, derived 
essentially from behavior theory, assumes that the fear of sorcerers 
occurs in societies where the direct expression of aggression is 
strongly inhibited and, hence, must be either attributed to others 
or justified by being directed against criminal sorcerers. The other 
view is derived from psychoanalytic theory and involves the hy- 
pothesis that sorcery implies paranoia, a personality variable which 
is derived from sexual inhibition and involves homosexuality. Whit- 
ing and Child (1953) were unable to decide between these two 
hypotheses. On the basis of their evidence, sorcery was found to 
be an important explanation for illness both in societies where chil- 
dren were punished severely either for sex or aggression during 
childhood. The fact that severity of socialization in these two be- 
havior systems are positively related to one another makes it difficult 
to disentangle their influence. 

Whiting (1959a) presents some evidence in favor of the sex 
anxiety hypothesis, but the data are not very convincing. The most 
likely interpretation of the results so far is that there are in effect 
two kinds of projection. The distinction between these may corre- 
spond to that which has been made between sorcery and witch- 
craft, the former being a result of the inhibition of aggression, the 
latter being associated with conflict in the area of sex. That sor- 
cerers are more often male and witches female is suggestive in this 

That aggression may be projected has been shown by Wright 
(1954) using a content analysis of folktales as an index. He showed 
that in societies with severe training in the control of aggression 


during childhood the hero in folktales does not direct his aggres- 
sion toward friends but rather toward strangers or enemies, that a 
stranger rather than the hero was more likely to be the agent of 
aggression, and finally that the hero was less likely to be triumphant. 
Whiting and Child (1953) report a similar finding. Societies with 
severe training in the control of aggression which believe that spirits 
can cause illness, tend to define the spirits as animal rather than 

The maintenance system variables relating to severe socialization 
for sex and aggression have already been reported — the former is 
associated with polygyny, the latter with the extended family 
household. Direct relationships between maintenance system vari- 
ables and sorcery were reported in two studies. Beatrice Whiting 
(1950) , assuming that sorcery functions as a mechanism of social 
control, showed that a strong belief in sorcery occurs in societies 
lacking in mechanisms of social control that involve the delegation 
of authority for the judging and punishing of crime. She also 
showed that this pattern tended to occur in small rather than in 
large societies. LeVine (i960) showed that sorcery tends to occur 
in societies that maximize jealousy between co-wives. In three East 
African societies similar in other respects, the preoccupation with 
sorcery was greatest among the Luo where co-wives lived in ad- 
jacent houses and virtually absent among the Kipsigis where the 
co-wives ordinarily live miles apart. He also reports that, cross-cul- 
turally, sorcery is a major cause of illness in 93 per cent of the so- 
cieties with polygynous households, 60 per cent of the societies with 
mother-child households, 5 3 per cent of the societies with extended 
family households, and only 3 6 per cent of the societies with nuclear 
households. The total pattern for predicting sorcery thus seems to 
be small societies with no formal systems of social control with 
either polygynous households and severe sex training or extended 
family households and severe training in the control of aggression. 

Independence Training and Achievement Motivation 

McClelland and Friedman (1952) report cross-cultural findings 
supporting the hypothesis that achievement motivation is produced 
by early and severe training in independence. Achievement motiva- 
tion was measured by applying to folktales a modification of the 
method used to score need achievement imagery in thematic apper- 
ception tests. Such scores were related to Whiting and Child's 
( 1953 ) measures of the age and severity of independence training. 


Societies with early and severe socialization of independence tended 
to have more achievement imagery in their folktales. 

Child, Storm, and Veroff (1958) also investigated the relation 
of child-rearing variables to achievement imagery in folktales. They 
used a larger sample of societies (the McClelland and Friedman 
study was restricted to North American tribes) and reported es- 
sentially negative results. Scoring reliability was low and different 
myth episodes from a single society showed wide variation in 
achievement imagery. They report the curious finding that socie- 
ties which are both generally severe in socialization and who punish 
achievement have more achievement imagery in their folktales than 
societies with any other combination of these child-rearing factors. 
They also report that positive training for achievement in later 
childhood is related to their folktale score if, and only if, training 
in self-reliance is held constant. Their score on achievement imagery 
was not significantly related to the Whiting and Child measures 
used in the McClelland and Friedman study. 

Over-All Early Soc/a//zaf/on, Decorative Art, and Asceticism 

Certain consequences are reported for over- all socialization anx- 
iety, a measure obtained by combining the scores for the five be- 
havior systems (Whiting and Child 1953). Barry (1957) reports 
that the decorative art forms of societies that are generally severe 
in training their children tend to be complex, and Friendly (1956) 
shows that such societies tended to have ascetic mourning customs. 
The relation of maintenance systems to over-all socialization anx- 
iety has not as yet been investigated. Fischer (1959) , however, re- 
ports that complexity in social structure is reflected in the com- 
plexity of decorative art. He used the Barry (1957) score on 
complexity of art design and Murdock (1957) scores of com- 
plexity of social organization. The presense of status distinctions 
based on wealth, social class membership, or heredity tends to result 
in complex designs in contrast with those from no rank distinctions 
or those based on age alone. 

Another over-all measure of the severity of socialization in early 
childhood is provided by the Barry, Bacon, and Child study ( 1957) . 
This measure, which they call "transition anxiety," is an estimate 
of the degree of pressure exerted upon the child during his change 
of status from infancy to childhood. This measure, although not 
statistically significant, is positively related to the Whiting and 
Child (1953) measure of over-all socialization anxiety. Whiting 


and his co-workers (ms.) show it to be related to household struc- 
ture. They report that societies with nuclear households are sig- 
nificantly more severe on this score than are societies with extended 
family households. It has already been pointed out that societies with 
nuclear family households begin independence training early. It 
now seems that they are generally severe as well, suggesting that 
strong pressures in child-rearing toward independence are required 
to enable a couple to set up an independent establishment. 

Socialization in Later Childhood 

An elaborate set of judgments about socialization during later 
childhood is provided by the Barry, Bacon, and Child study ( 1957) • 
These judgments concern the manner in which a child is trained to 
be obedient, responsible, self-reliant, nurturant, and generally in- 
dependent, as well as his training in achievement. For each of these 
behavior systems a separate judgment was made for the general 
pressure exerted upon the child, the severity of punishment for non- 
compliance, the difficulty of performance, the amount of conflict, 
and the frequency of the response. 

Separate judgments on the above scales were made for the treat- 
ment of boys and girls by Barry, Bacon, and Child (1957). Sig- 
nificant differences in training were reported. These involved more 
stress upon nurturance, obedience, and responsibility for the girls 
and upon achievement and self-reliance for the boys. Although 
they did not relate these differences to any projective system, they 
did report that large differences in the training of the sexes occur 
in societies where large animals are hunted, where grain rather than 
root crops are grown, where large or milking animals are kept, 
where fishing is unimportant or absent, where the settlement is 
nomadic rather than sedentary, and where polygyny is high. They 
interpreted those results as implying that differential training for 
boys and girls is required where superior strength and motor skill 
is involved or where a large family with a high degree of co-opera- 
tion is required. 

Lambert, Triandis, and Wolf ( 1959) , in the study discussed pre- 
viously (see page 357) concerning the nature of the gods, report 
that the supernatural are more aggressive in societies which put 
strong pressure upon the boys for self-reliance and independence. 
They also report an even stronger relationship in the same direction 
with a score which combines the pressures exerted in all six systems; 
that is, nurturance, obedience, self-reliance, achievement, responsi- 


bility, and general independence. It is interesting that they assume 
a reverse direction of causation to explain this relationship; that is, 
the belief in aggressive gods requires training a child to be inde- 
pendent and self-reliant so that he can cope with a hostile world as 
an adult. 

Bacon, Child, and Barry (ms.) show that societies which severely 
punish their older children for disobedience, irresponsibility, lack of 
self-reliance, and lack of achievement are high in the frequency of 
theft. Since they also find that a high frequency of theft is found in 
societies with low infant indulgence and severe weaning, they in- 
terpret these findings as a reaction to emotional deprivation during 
infancy and childhood. Such anxieties, except for severe weaning, 
interestingly enough, are not related to the frequency of personal 

Barry, Child, and Bacon (1959) report some interesting relation- 
ships between socialization pressures in later childhood and various 
aspects of the maintenance system, in this case the basic economy. 
They state, "In considering the relation of economy to adult role, 
and hence to child training, we felt that perhaps a variable of great 
significance is the extent to which food is accumulated and must 
be cared for," To test this hypothesis they classified societies into 
four categories on the basis of their subsistence activities which rep- 
resent the degree to which this implies an accumulation of food. 
Assuming that food "on the hoof" requires the greatest amount of 
care, societies that were mainly dependent upon animal husbandry 
were judged to be highest on the scale. The lowest point was rep- 
resented by hunting and fishing societies. Between these extremes a 
distinction was made between those societies depending upon agri- 
culture only for subsistence and those depending upon a combina- 
tion of agriculture, hunting, and fishing. The former were assumed 
to be higher in food accumulation than the latter. 

Contrasting the extremes on the scale, that is, animal husbandry 
versus hunting and fishing, they showed that societies with high 
accumulation of food put strong pressure upon their children to be 
responsible and obedient and were low in stressing achievement and 
independence in their boys and also low in stressing achievement 
and self-reliance in girls. They then constructed a general score 
which they called "pressure toward compliance versus assertion" 
by adding the scores on obedience and responsibility and subtract- 
ing from this sum the combined score on achievement and self- 
reliance. The relation of this over-all pressure toward compliance 




Subsistence Economy 


on Compliance 


Animal Husbandry 



Agriculture only 



Agriculture, hunting 

and fishing 



Hunting and fishing 



Table 2. Relation between pressure toward compliance versus assertive- 
ness as indicated by a subsistence economy scale. Numbers in parentheses 
represent the number of societies in each category. This table is adapted from 
Barry, Child, and Bacon 1959, P. 60. 

to food accumulation is striking and shown in Table 2. 

It should be noted that high points on the subsistence scale — ani- 
mal husbandry and agriculture — are rather heavily weighted with 
cases from Africa and that perhaps pressure toward compliance is 
an African culture trait and, thus, the association is spurious. If, 
however, all African cases are omitted from the sample, the asso- 
ciation between subsistence and pressure toward compliance is still 
strong. When this is done, high compliance is represented by the 
following percentages in order of the degree of accumulation: 70 
per cent, 90 per cent, 33 per cent, and 14 per cent. Thus, the re- 
lationship, although somewhat less strong, is still substantial. 

Although the direct relationship between subsistence economy 
and aggressive gods is not reported, Bacon, Child, and Barry (ms.) 
indicate that this scale is not related to the frequency of theft. Thus, 
here again a child-rearing factor seems to be a necessary link be- 
tween an aspect of the maintenance system and a projective con- 
sequence. It should be noted that D'Andrade in this volume, using 
a scale for measuring subsistence economy essentially similar to the 
one described above, found this aspect of the maintenance system 
directly predicts a projective measure — a preoccupation with 
dreams — but that neither were related to child rearing. 


The general hypothesis that personality can serve as a mediator 
between the maintenance and projective systems of a culture has 
been supported by a fairly substantial amount of cross-cultural re- 
search. It should be pointed out, however, that a substantial num- 
ber of specific hypotheses failed to be confirmed. Many of these 
have not been reported in the studies under review; furthermore, 
those negative findings which were reported have usually been 
omitted from this review. To have included them would have been 


too cumbersome. This decision, however, may give an exaggerated 
view of the importance as an integrating factor of those personahty 
processes which are determined by child rearing. It should be noted 
that both Wallace and D'Andrade report in this volume cultural 
responses relevant to personality which are not related to child 
rearing, but rather to either physiological process or to social struc- 
ture. Despite these cautions, it seems clear that economics and social 
structure do often have a determining influence upon the way in 
which children are brought up, and the child rearing in turn often 
has a predictable and determining effect upon magical belief, rit- 
uals, art forms, taboos, and even crime rates. 

The direction of causation is, of course, an ever-present problem 
in cross-cultural research. Why cannot it be assumed that projective 
systems determine child rearing and that child rearing determines 
the maintenance systems? Such may, in fact, be the case in some or 
even many of the instances reported above. Although this is an 
important question, it is beyond the scope of this review. The posi- 
tion taken here has been to accept the assumptions as to the direc- 
tion of causation made by the authors of the works considered. 

In reviewing the relationships reported above, certain patterns 
or types emerge which should be noted. First, where a certain aspect 
of the maintenance system may be classed into discreet categories, 
such as household, marriage form, residence, basic subsistence econ- 
omy, and so forth, some of these categories may be determining 
with respect to child rearing, whereas other categories in the same 
maintenance system may be nondetermining. A number of exam- 
ples of this contrast could be drawn from the results reported above, 
but the relation between household structure and over-all infant in- 
dulgence shown in Table i will serve to illustrate this contrast. It 
will be seen from this table that extended and polygynous family 
households determine high infant indulgence but that nuclear 
households are nondetermining as to indulgence. Mother-child 
households are, by a confidence limits test (Hald 1952), almost 
determining of low indulgence, but the proportion does not quite 
reach the 5 per cent level of confidence. Thus, if a person were told 
that a society had an extended family or polygynous household, 
he could make money, even if he gave odds, that infants were 
treated indulgently, but if he were told that a society had nuclear 
households, he should not bet on how infants are treated. It is be- 
lieved that this distinction between cultural categories which are 
determining and those which are nondetermining may apply to 


Other features of cultural integration than child rearing and may 
be a distinction useful to keep in mind in describing cultural pat- 

A second general conclusion may be drawn from this review. 
This consists of a typology of various ways in which personality 
factors can serve to integrate culture. These types can perhaps best 
be shown in the following diagrammatic models. In these diagrams 
M will stand for a maintenance system variable, C for a child-rear- 
ing practice presumed to influence personality, and P for a projec- 
tive system variable. The arrows stand for the assumed direction of 

The most common type is shown in Figure i . Here it is assumed 

M P 


Figure 1. The mediation type. 

that a certain feature of maintenance systems determines a child- 
training practice and that this practice determines a feature in the 
projective systems, but that the given feature of the maintenance 
systems has no directly determining influence with respect to the 
projective system feature. This type can be illustrated by the rela- 
tion between household structure, infant indulgence, and the na- 
ture of the gods as described above: that is, extended family house- 
holds predict high infant indulgence, and high infant indulgence 
predicts a low fear of ghosts, but household structure is unrelated 
to the fear of ghosts (see p. 360) . 

Another important type is shown in Figure 2. Here it is assumed 
that neither a feature of the maintenance system nor a child-rearing 
practice alone will determine a given feature in the projective sys- 
tem but that taken together they will. As an example, the age of 
weaning predicts guilt in monogamous but not polygynous societies 
(see p. 370). 




Figure 2. The interaction type. 

It is more likely that many more examples of this type will be dis- 
covered as research in this area becomes more sophisticated. 


The third type, shown in Figure 3, assumes a direct effect of pres- 
sures from the maintenance system upon some aspect of the projec- 
tive system. 

M >F 


Figure 3. Adult pressure type. 

Although this type was not considered in this review, a recent study 
(Field i960) showing heavy drinking to be associated with bilateral 
descent, but to none of the child-rearing variables discussed above, 
is a good example of this type. 

Figure 4 indicates the assumption of causation between a child- 
rearing and a projective feature in a direction opposite to that which 

M P 


Figure 4. "Reverse" Causation. 

has usually been assumed in the studies under review. The only 
case of this type noted is from the study by Lambert and his group 
(1959) where training in self-reliance and independence was inter- 
preted as being a consequence rather than the cause of a belief in 
aggressive gods (see p. 357). 

As a final comment it would seem to this reviewer that the cross- 
cultural study of personality as a mediating factor in the integration 
of culture is off to a good start, but still has a long way to go. The 
measurement of child rearing is far from satisfactory partly because 
ethnographic reports are often inadequate and partly because it is 
highly unlikely that the variables selected by Whiting and Child 
(1953) and by Barry, Bacon, and Child (1957) will turn out in 
the long run to be more than first approximations of the dimen- 
sions most crucial to personality development. Furthermore, as has 
been shown in this review, cross-cultural research has just begun to 
attack the complex problem of the effects of the interaction of sev- 
eral variables operating jointly — an approach which should yield 
interesting results in the near future if it is pursued. 



Ayres, Barbara C. 

1954 Personality determinants of food and sex taboos during pregnancy. 
(Typescript, Doctoral dissertation.) Cambridge, Radcliffe College. 

Anthony, Albert S. 

1955 A cross-cultural study of factors relating to male initiation rites and 
genital operations. (Typescript, doctoral thesis. Harvard Graduate 
School of Education.) Cambridge. 

Bacon, Margaret K., I. L. Child, and Herbert Barry, III 

A cross-cultural study of crime in pre-literate societies. (Typescript). 

Barry, Herbert A. 

1957 Relationships between child training and the pictorial arts. Journal 
of Abnormal and Social Psychology 54:380-383. 

Barry, Herbert A., Margaret K. Bacon, and Irvin L. Child 

1957 A cross-cultural survey of some sex differences in socialization. Journal 
of Abnormal and Social Psychology 55:327—332. 

Barry, Herbert A., Irvin L. Child, and Margaret K. Bacon 

1959 Relation of child training to subsistence economy. American Anthro- 
pologist 61:1. 
Benedict, Ruth 

1934 Patterns of culture. Boston, Houghton-Mifflin Co. 
Blalock, H. M., Jr. 

i960 Correlational analysis and causal inferences. American Anthropologist 

Burton, Roger V. and J. W. M. Whiting 

i960 The absent father: effects on the developing child. Paper presented at 
A. P. A. Meeting, September, i960. 

Child, Irvin L., T. Storm, and J. Veroff 

1958 Achievement themes in folktales related to socialization practices. In 
Motives in fantasy, action and society, J. W. Atkinson, ed. New York, 
Van Nostrand. 

Field, Peter B. 

A new cross-cultural study of drunkenness. In Society, culture and 
drinking patterns, D. J. Pittman and C. R. Snyder, eds. New York, 
Wiley. In press. 

Fischer, John L. 

1959 Art styles and cultural cognitive maps. Paper presented at American 
Anthropological Association Meeting, Mexico City, December, 1959. 

Friendly, Joan P. 

1956 A cross-cultural study of ascetic mourning behavior. (Typescript, 
Honors Thesis) Cambridge, Radcliffe College. 

Ford, Clellan S. 

1945 A comparative study of human reproduction, Yale University Publica- 
tions in Anthropology, No. 32. New Haven. 


Hald, a. 

1952 Statistical tables and formulas. New York, Wiley. 
Kardiner, Abram and Ralph Linton 

1939 The individual and his society. New York, Columbia University Press. 
Kardiner, Abram et al. 

1945 The psychological frontiers of society. New York, Columbia University 

Lambert, W. W., Leigh Triandis, and Margery "Wolf 

1959 Some correlates of beliefs in the malevolence and benevolence of 
supernatural beings: A cross-cultural study. Journal of Abnormal and 
Social Psychology 58:2. 

LeVine, Robert A. 

i960 Witchcraft and marital relations in East Africa: a controlled compar- 
ison. Paper presented at the American Anthropological Association 
Meeting, Minneapolis, Minnesota. 

McClelland, D. C. and G. A. Friedman 

1952 A cross-cultural study of the relationship between child-training prac- 
tices and achievement motivation appearing in folk tales. In Readings in 
social psychology (rev. ed.), G. E. Swanson, T. M. Newcomb, and 
E. H. Hartley, eds., pp. 243-249. New York, Henry Holt. 
Mead, Margaret 

1928 Coming of age in Samoa. New York, Morrow. 

1935 Sex and temperament in three primitive societies. New York, Morrow. 
MuRDocK, George P. 

1957 World ethnographic sample. American Anthropologist 59:664-687. 
MuRDOCK, George P. and J. W. M. Whiting 

195 1 Cultural determination of parental attitudes: the relationship between 
the social structure, particularly family structure and parental behavior. 
In Problems of infancy and childhood, Milton J. E. Senn, ed. New York, 
Josiah Macy, Jr., Foundation. 

Prothro, E. Terry 

i960 Patterns of permissiveness among preliterate peoples. Journal of Ab- 
normal and Social Psychology 61:151—154. 

Roberts, John M., Robert R. Bush, and Malcolm Arth. 

1957 Dimensions of mastery in games. Stanford, California, Ford Center for 
Advanced Study in the Behavioral Sciences. (Mimeographed.) 

Spiro, Melford E. and Roy G. D'Andrade 

195 8 A cross-cultural study of some supernatural beliefs. American Anthro- 
pologist 60:456—466. 

Stephens, William N. 

The oedipus complex: cross-cultural evidence. Glencoe, Illinois, The 
Free Press. (In press.) 

Whiting, Beatrice B. 

1950 Paiute sorcery. Viking Fund Publications in Anthropology No. 15, 
New York. 


Whiting, Marjorie Grant 

1958 A cross-cultural nutrition survey of 118 societies representing the 
major culture areas of the world. Unpublished Ph.D. thesis, Harvard 
School of Public Health. 
Whiting, J. W. M. 

1959a Sorcery, sin and the superego: a cross-cultural study of some mechanisms 
of social control. In Symposium on motivation, pp. 174—195. University 
of Nebraska Press. 

1959b Cultural and sociological influences on development. In Maryland child 
growth and development institute, June 1-5, 1959, pp. 5—9. 

1959c The male and female conscience. Paper presented at American Psycho- 
logical Association Meeting, September. Cincinnati. 

1960a Social structure and identification. Mona Bronfman Sheckman Lectures 
delivered at Tulane University, New Orleans, Louisiana. 

1960b Resource mediation and learning by identification. In Personality de- 
velopment in children, L Iscoe and M. Stevenson, eds. Austin, Texas, 
University of Texas Press. 
Whiting, J. W. M., M. F. Antonovsky, E. M. Chasdi, and B. C. Ayres 

The learning of values. In Peoples of Rimrock (Vol. L), E. Z. Vogt 
and J. M. Roberts, eds. Final Report of the Harvard Values Study, 

Whiting, J. W. M. and Irvin L. Child 

1953 Child training and personality. New Haven, Connecticut, Yale Uni- 
versity Press. 

Whiting, J. W. M., Richard Kluckhohn, and Albert S. Anthony 

1958 The function of male initiation ceremonies at puberty. In Readings in 
social psychology, Eleanor E. Maccoby, T. Newcomb, and E. Hartley, 
eds., pp. 359-370. New York, Henry Holt, 

Wright, George O. 

1954 Projection and displacement: a cross-cultural study of folk-tale ag- 
gression. Journal of Abnormal and Social Psychology 49:523-528. ■ 

chapter i^ 



Brand e is University 

Historical Perspeciive 

It would be fair to say that in the field of culture and personality, 
child-rearing practices have been studied primarily from the point 
of view of their effects on the development of personality, rather 
than as products of other features of the culture (cf. Child 1954) . 
Sometimes the inquiry about the effects of socialization stops with 
the attempt to demonstrate the impact of child rearing on person- 
ality. Sometimes it goes on to attempt to show that various cultural 
features are derived from imputed or known personality character- 
istics present in a given group. In either case, the child-rearing prac- 
tices are viewed as causes, and personality or features of culture as 

* This essay owes a great deal to a series of graduate seminars on the topic of the causes of 
socialization which I have conducted at the University of Michigan from 1954 to the present. 
I am indebted to many students for discussions, research results, and papers on theoretical prob- 
lems — so many that it would be invidious to select a few for mention and impossible to list them 
all. I am grateful to the Social Science Research Council for funds which provided me with a 
research assistant for one of these seminars, and to Mrs. Eviva Menkes for her able work in that 
capacity. I benefited by several years' stimulating meetings of a Social Science Research Council 
Committee on Personality Development, the other members of which were Alfred L. Baldwin, 
William E. Henry, Robert R. Sears, M. Brewster Smith (stafF), and John W. M. Whiting. An 
earlier version of this essay was prepared for a conference on cross-cultural research on personality 
development sponsored by the SSRC committee just mentioned and held in Kansas City on May 
20—22, 1955. I profited by discussions of that earlier version at the conference. I am indebted for 
helpful criticism and discussion to John W. Atkinson, Thomas Gladwin, my wife, E. Kathleen 
Gough, Francis L. K. Hsu, Alex Inkeles, the late Clyde Kluckhohn, Robert LeVine, Daniel R. 
Miller, Kaspar D. Naegele, and G. E. Swanson, some of whom have read one or another draft of 
this paper, and some of whom have discussed the general problem with me. The University of 
Michigan provided travel funds which made it possible for me to attend a conference with the 
editor of this volume and some of the authors of other chapters. 

I am especially grateful to Irvin L. Child, who provided me with ratings on the socialization 
practices of 1 1 1 cultures, prepared by him and his co-workers, and as yet unpublished. These 
ratings have been used in many of the University of Michigan seminars mentioned above. Some 
results of this work are mentioned below. 



Less effort has been made to determine the causes of the socializa- 
tion practices themselves. Sometimes these causes are treated as 
self-evident; sometimes the problem is disregarded. This essay at- 
tempts to set forth an approach oriented to systematic inquiry into 
the causes of socialization patterns. It will discuss the theoretical 
utility of such an approach, and will outline some of the features 
of cultural systems which seem to be important causes of socializa- 
tion patterns. It will make some mention of field techniques and 
comparative techniques and will allude to some results now avail- 

The Problem 

Anthropologists have been willing to treat joking and respect 
relationships as the outgrowth of other features of kinship relation- 
ships (Radcliffe-Brown 1952; Eggan 1955), to examine kinship 
terms as reflexes of kinship groupings (Murdock 1949) , to see po- 
litical structure and social complexity as functions of level of pro- 
ductivity (White 1949) , but by and large they have not been in- 
terested in accounting for socialization practices. 

There are points of view in culture and personality which make 
attention to the causes of socialization seem unnecessary or un- 
profitable. One of these sees socialization as the prime cause of major 
features of different cultural systems, but pays no heed to the ques- 
tion of why the members of a particular system show uniformities 
in socialization, rather than randomness. A second point of view, 
which confines itself to relations between socialization and per- 
sonality, is one or another version of the "chicken-and-egg" ap- 
proach. In its simplest form, this theory would hold that people who 
grow up under a given socialization regime reproduce the same 
regime that they experienced, because the personalities they de- 
veloped make it congenial to do so. In this version, antecedent and 
consequent pursue each other in a small circle forever, and the an- 
swer as to why the socialization pattern is as it is can only be because 
the socializers were reared as they were. There is no room in this 
system for change. A more sophisticated chicken-and-egg approach 
asserts that strains engendered under one socialization regime give 
rise to efforts by those who experienced the strains to alter the 
regime in rearing their own children. Presumably the result is either 
a new stability or a perpetual series of changes, but in this version, 
too, socialization and personality chase each other forever. 

When we find, however, as Child and his co-workers have done, 


in studies discussed below, that there is sizable variation in socializa- 
tion aims in different types of subsistence economies, none of these 
views seems particularly satisfactory. For now we see that the eco-